Chapter 19

Nonautomotive Diesel Electric Locomotives, Appliances, Aviation

General Motors manufactures not only cars and trucks, but diesel electric locomotives, household appliances, aviation engines, earth-moving equipment, and a variety of other durable goods; altogether, our nonautomotive business accounts for roughly 10 per cent of our civilian sales. And yet there have always been limits to our product diversification. We have never made anything except "durable products," and they have always, with minor exceptions, been connected with motors. Not even Mr. Durant, for all his expansion and diversification, ever suggested that we should stray into any broad field clearly outside the boundary suggested by our corporate name, General Motors.

No attempt will be made here to present detailed individual histories of our products outside of the automobile. The stories of our pioneering in the diesel business, of our development of the Frigidaire line of products, and of our aviation business are the subjects of this chapter.

It would be nice to be able to trace a coherent pattern in General Motors' ventures outside the automobile business, but chance and other factors that entered the picture make it difficult to do so. We had, of course, some natural interest in diversification which might afford us a hedge against any decline in automobile sales. But we never had a master plan for nonautomotive ventures; we got into them for different reasons, and we were very lucky at some crucial points. We got into the diesel field, for example, because of Mr. Kettering's special interest in diesel engines, dating back as early as 1913, when he was experimenting with diesel power in an attempt to find a suitable engine for the generator in a farm-lighting set he wanted to manufacture. Mr. Durant put General Motors in the refrigerator business for reasons of his own; but it is clear, as I shall show, that we would have abandoned Frigidaire in its earliest years had it not been for an odd combination of events. And we got into aviation because we thought the small airplane would be an important competitor of the automobile.

It is worthy of note, I believe, that these were relatively new products at the time we first invested in them. There was no diesel locomotive capable of providing mainline service on American railroads; the electric refrigerator was only an impractical gadget, and the future of aviation was anybody's guess. In other words, we did not simply use our financial and engineering resources to "take over" new products outside the automobile business. We got in early—as long as forty-five years ago—and helped develop them. Our operations in these fields have been expanded, but we have gone into nothing entirely new in more recent years except for the purchase in 1953 of the Euclid Road Machinery Company (manufacturer of earth-moving equipment), and war and defense production.

Diesel Electric Locomotives

General Motors entered the locomotive industry in a small way in the early 1930s. At the time, railroads in the United States seemed to have very little interest in diesel locomotives except for special switch-engine use. Yet in less than a decade the diesel was outselling the steam locomotive, and General Motors was outselling all other locomotive manufacturers combined. Because we led the diesel revolution, with tremendous savings to the railroad industry, the Electro-Motive Division today enjoys a large part of the locomotive market.

There were, I think, two principal reasons for this rather spectacular progress. The first was simply that we were more tenacious in our efforts to produce lightweight, high-speed diesel engines suitable for over-the-road use on American railroads. The second reason was that we brought to the locomotive industry some of the manufacturing, engineering, and marketing concepts of the automobile industry. Until we began making diesels, locomotives had always been produced on a custom basis, with the railroads specifying their requirements to the manufacturers in considerable detail so that virtually no two locomotives on American railroads were alike. But almost from the beginning we offered the railroads a standard locomotive—one that we were able to produce in volume at a relatively low price. In addition we guaranteed performance at a lower net cost per ton mile than was possible with the use of steam engines, and we made good our guarantees by maintaining a service organization and providing standardized replacement parts. This program revolutionized the locomotive industry and secured our own place in it.

There was, of course, nothing new about the principle of the diesel engine at the time that General Motors first became interested in it. Rudolph Diesel, a German inventor, received the original patent for this kind of engine in 1892 and built a successful unit with one cylinder and twenty-five horsepower in 1897. As early as 1898 a sixty-horsepower, two-cylinder diesel unit was built in this country. These early devices embodied essentially the same compression-ignition principle as the engine in a modern diesel locomotive.

The four-cycle diesel engine works this way: On the first suction stroke of the piston, the engine draws in air and nothing else. The next stroke of the piston compresses the air to something like 500 pounds per square inch, creating a temperature of around 1000 Fahrenheit. Just before the end of the compression stroke, oil is injected as a fine spray into the combustion chamber under high pressure. The hot air ignites this fuel. The third and fourth strokes of the piston provide the power and exhaust—as in a gasoline engine. However, the diesel requires neither a carburetor nor an electric ignition, and thus has an edge in simplicity over the gasoline engine.

As this description indicates, the diesel converts its fuel directly into a source of energy. In this respect it is unlike the steam engine, whose fuel is used only to create steam, and unlike the gasoline engine, which vaporizes its fuel before it can be ignited. Both of these engines are less efficient than the diesel—which has in fact the highest thermal efficiency of any heat engine in everyday use. The modern diesel uses a distilled petroleum fuel oil, but other fuels have been used in the past. Rudolph Diesel himself had intended to run his engine on powdered coal, but his engineer associates persuaded him at the outset to use petroleum oil in order to avoid the problem of scoring. Powdered coal was used later experimentally by others attempting to follow Diesel's original intentions, and other fuels have been tried. But petroleum oil remains the standard diesel fuel.

Despite its great efficiency, the diesel engine was for many years quite limited in practical use. With few exceptions the engines were large, heavy, and slow running, and so found their greatest application in power stations, pumping, and marine use. They weighed 200 or 300 pounds per horsepower, and this, indeed, was the heart of the problem—to build a powerful, fast-running diesel of relatively small size.

I have said there was nothing new about the principle of the diesel engine. I might add that there were no unknown principles concerning any component part of the diesel-powered locomotive that General Motors created. What was lacking was the imagination, the initiative, and the talent to work out the problem to the point of practicability.

Europeans had been working on this development since the second decade of this century and had some diesel railcars and locomotives in operation by 1920. By 1933 a few U.S. diesel manufacturers had successfully built a number of diesel engines for switcher service. Since weight was an advantage in switchers, and since they showed economies over steam in operation, they met with some success. However, attempts to build diesel engines for mainline passenger and freight applications in this country were not successful, since in these cases weight, power, and size are critical. Bringing the diesel engine down to more manageable proportions, with a low weight-per-horsepower ratio, was the principal concern of our engineers.

In a large organization like General Motors it is seldom possible to assign to any one person the credit—or blame—for initiating some major undertaking. But in the case of the diesel, Charles F. Kettering comes very close to being the whole story. The General Motors Research Corporation, forerunner of our present Research Laboratories, was testing diesel engines, under Mr. Kettering's close scrutiny, as early as 1921. After Mr. Kettering bought himself a diesel-powered yacht in April 1928 these engines became a major preoccupation of his. As anyone who knew him might have guessed, when on his yacht he was more often tinkering in the engine room than relaxing on deck. He was already convinced that the diesel did not have to be unreasonably large and heavy.

I became interested in the possible development of the diesel engine for General Motors at about the same time. If my memory serves me correctly, I remember dropping in one day at the Research Laboratories in Detroit and saying to Mr. Kettering: "Ket, why is it, recognizing the high efficiency of the diesel cycle, that it has never been more generally used?" In his characteristic way he said the reason was that the engine would not run in the way that the engineers wanted it to run. I then said to him: "Very well—we are now in the diesel engine business. You tell us how the engine should run and I will see that available manufacturing facilities are provided to capitalize the program." Of course, saying that we were in the diesel business was a manner of speaking. I meant I would support him in the organization.

In 1928 Mr. Kettering and an engineering group at the Research Laboratories began a series of comprehensive tests on the diesel engines then being offered by various manufacturers. An analysis of these tests, combined with a thorough study of current scientific literature on diesels, finally led Mr. Kettering to conclude that the solution to his problem was the so-called two-cycle diesel engine. The two-cycle engine was nothing new at that time. Indeed, the truly remarkable feature of Mr. Kettering's conclusion was his conviction that the two-cycle principle was ideally suited to the smaller diesel engines. Though it had been thoroughly explored before, it generally had been rejected as unworkable except in large, slow speed engines.

In the two-cycle engine the intake of fresh air and the exhaust of burned gases take place at the same time. One stroke out of every two is a power stroke instead of one out of four, as in the four-cycle engine. The result is an engine that has less than one fifth the weight and one sixth the size of its predecessor four-cycle engine of equivalent power output. But this smaller device created some awesome engineering problems. For one thing, the two-cycle engine as developed by Mr. Kettering called for much greater precision in the fuel-injection system. Specifically, what the Research Laboratories were called upon to produce—and finally did produce —was a unit fuel injector whose parts fitted with a clearance of 30 to 60 millionths of an inch and an injector pump which built up pressures as high as 30,000 pounds per square inch as it forced fuel through holes 10 to 13 thousandths of an inch in diameter drilled in the injector tip. The two-cycle engine also has to have an external air pump. This became another major project, but finally Research delivered what was needed: a light, compact device able to pump large quantities of air at a pressure of about three to six pounds.

By the end of 1930 it was clear that the two-cycle engine was practical and that Mr. Kettering had achieved a major breakthrough in diesel technology. It was also clear that the time had come to provide the manufacturing facilities I had promised him. We looked around for the special facilities that were needed. Our buildup consisted principally of the purchase of two companies: the Winton Engine Company and the Electro-Motive Engineering Company, both of Cleveland, Ohio.

Winton was a manufacturer of diesel engines, primarily for marine uses (it had built Mr. Kettering's second set of yacht engines), and also of certain kinds of large gasoline engines. ElectroMotive was an engineering, design, and sales firm with no manufacturing facilities of its own. The two firms had had an intimate business relationship for almost a decade. During that time ElectroMotive and Winton had built a substantial business and reputation in the design and sale of gas-electric railcars, primarily for use on short-haul runs. Building the engines for these railcars was a major part of Winton's business during most of the 1920s. Relative to steam, however, the operating economies of gas cars kept diminishing, and toward the end of the decade Electro-Motive began to find itself in trouble trying to continue to sell the gas-electric car, which in turn had its effect on Winton.

Against this background Winton and Electro-Motive began, around 1928 and 1929, to look seriously into the possibility of using diesel power on the railroads. Harold Hamilton, then president of Electro-Motive, encountered the same problems of fuel injection that Mr. Kettering was then wrestling with. Mr. Hamilton was also trying to develop a small diesel engine. With the technology then available to him, the smallest diesel he could build was one weighing about sixty pounds per horsepower. A locomotive, he felt, required an engine weighing no more than twenty pounds per horsepower, with a crankshaft speed of about 800 revolutions per minute. Though there were a few available diesel engines which closely matched these specifications, Mr. Hamilton did not feel that they could stand up to the performance and reliability requirements which he felt were necessary for successful railroad application. Furthermore, Mr. Hamilton realized that the diesel he wanted would require metal tubes and joints able to last for long periods of time even when they had to carry fuel under pressures of 6000 and 7000 pounds to the square inch. Winton was not able to develop this kind of metallurgy and Mr. Hamilton knew of no place in the industry where it was available. He finally concluded that it would take about $10 million of venture capital to solve his and Winton's problems—perhaps $5 million to overcome the technological obstacles, and another $5 million or thereabouts to provide the plant and equipment needed for manufacturing facilities.

It was speedily made apparent to Mr. Hamilton, and also to George W. Codrington, the president of Winton, that they would not be able to raise the money at the banks and that there was certainly no such venture capital anywhere in the railroad industry. (Neither the carriers nor the locomotive manufacturers showed enough interest in the diesel to undertake the research necessary.) At about this time, however, Mr. Kettering became acquainted with Mr. Codrington as the result of ordering Winton engines for his second yacht. He bought these engines simply because Codrington agreed, though reluctantly, to put in a new kind of injector that one of the Winton engineers was developing at the time, and which Mr. Kettering felt held great promise. I don't know who first suggested the idea of Winton's coming into General Motors. In any event, we began to negotiate formally in the late summer of 1929. Agreement on the purchase of Winton had almost been reached in October when the great market crash temporarily confused the picture.

But there was never any serious question in our minds that Winton was a good buy for us. For one thing, we were not at this point certain about the future of the U.S. automobile market, which had not been expanding during the late 1920s. Consequently, we had a natural interest in any enterprise within our scope that offered us a reasonable opportunity to diversify.

The case for buying Winton was stated by John L. Pratt, vice president, in a memorandum which he addressed to the Operations and Finance committees, dated October 21, 1929, as follows:

We have had under consideration for some time past the possible purchase of the Winton Engine Company located at Cleveland, Ohio, which subject has been informally discussed at previous meetings.

It is believed that the Diesel Engine development in this country has arrived at a point where it has become commercial and is probably on the eve of considerable expansion. The Winton Engine Company is unquestionably the outstanding Diesel engine manufacturer in the United States . . .

The Winton Company has a capable management and would not require any additional personnel immediately. If the business continues to expand, as we believe it will, we may think it desirable to add to its personnel another good executive, perhaps as Assistant General Manager or Sales Manager . . .

. . . The purchase of this company will give us a vehicle for capitalizing the developments of our research organization along engine lines and will assist materially in keeping us abreast of Diesel engine developments. The business should also be reasonably profitable, and if expansion continues, as most of our engineers believe it will, we should ultimately make a good return on the investment required to purchase the Winton Company . . .

Finally, in June 1930 the Winton operations became a part of General Motors with Mr. Codrington continuing as president. Winton's principal market continued to be in large marine engines. (Note 19-1.)

Five months after the Winton acquisition we also acquired ElectroMotive and again the old management of the company continued to run its affairs. During the negotiation to acquire Electro-Motive, Mr. Hamilton and Mr. Kettering continued to hold many lengthy discussions about the challenge of the lightweight diesel engine. In his 1955 testimony before a Senate subcommittee Mr. Hamilton described the tremendous enthusiasm of Mr. Kettering for the job of developing a diesel engine: ". . . it was just like ringing a bell to a fire horse," he recalled. Mr. Hamilton, in fact, made it clear that he was not attracted to General Motors merely by the corporation's great economic strength. ". . . we had more than that in General Motors," he commented. ". . . of the companies that I knew at that time, many of them with plenty of financial resources, none of them had the mental approach to this problem that was necessary to take it at that stage that it was in then, and the courage that went along with it to move it to its point of success. At least that was our opinion in the matter."

For a while Winton and Electro-Motive operated about as before. Mr. Hamilton and Mr. Kettering both had the impression that it would take a considerable length of time to build a commercially acceptable diesel engine for the railroads. Meanwhile, Mr. Kettering devoted his efforts, in large measure, to perfecting the two-cycle diesel engine. By 1932 Mr. Kettering decided he could build a two-cycle, eight-cylinder engine that would produce about 600 horsepower. Since Mr. Kettering's new engine would have a good edge over existing four-cycle engines in the 600-horsepower range, particularly in the weight-per-horsepower ratio, his engine seemed worth building.

At about this time we were planning our exhibit for the Century of Progress World's Fair, which was scheduled to open in Chicago in 1933. Our exhibit was to be a dramatic display—an automobile assembly line in actual operation, producing Chevrolet passenger cars. We needed a source of power for the assembly line and decided that two of Mr. Kettering's proposed 600-horsepower diesel engines would do the job.

When we first conceived the idea of powering our World's Fair display with the new diesel engine, what we had in mind was to get a good long look at the engine under actual operating conditions. We were primarily concerned with proving that Mr. Kettering's basic design was a good and practical one; we did not anticipate that the commercial applications would come as soon as they did. But before the engine for the display was even finished, our perspective on this matter was drastically altered.

What changed it, principally, was the sudden interest of one railroad president—Ralph Budd of the Burlington—in the diesel engine. Mr. Budd was then hoping to build a new, streamlined, lightweight passenger train that would be dramatic in appearance and economical in operation. One day in the fall of 1932 he stopped off in Cleveland to see Mr. Hamilton, who told him about General Motors' diesel experiments and put him in touch with Mr. Kettering. Mr. Budd was excited about the prospects.

He paid a visit to Detroit and to the General Motors Research Laboratories. Mr. Kettering showed him the experimental two-cycle engine but warned him that the eight-cylinder model was not yet built and certainly required a great deal more development work before it could be considered seriously as a source of locomotive power. Mr. Budd was told about General Motors' plans to test the engine at the World's Fair.

When the fair finally opened, our diesel engines were visible, through a plate-glass window, to anyone who cared to inspect them. However, we were still apprehensive about them, and the publicity man for our exhibit was under strict orders to say nothing about them—even though they were, in a sense, the most dramatic feature of our exhibit. The engines were unheralded, then, but Mr. Budd, at least, paid close attention to them during the entire fair. He was well aware of the difficulties we were having with the engines. He knew that every night one or two engineers had to work on them to ensure that they would still be functioning the next day. He knew the opinion of Mr. Kettering's son, Eugene, who was in charge of the maintenance operation, and who later commented that "the only part of that engine that worked well was the dipstick."

Nevertheless, Mr. Budd continued to press us for a diesel engine that he could use on his Burlington Zephyr. He became more insistent than ever when, in 1933, the Union Pacific publicly announced its plans to build a streamlined train. The Union Pacific was planning only a small, three-car affair without any real locomotive—the power car was to be an integral part of the train itself. The power was derived from a twelve-cylinder, 600-horsepower gasoline engine, which was built by Winton. There were no major technological innovations in this Union Pacific train; but pictures of it were widely distributed, the public reception was quite favorable, and suddenly the nation was very much interested in streamliners. All of this served to fortify Mr. Budd's desire, which was intense anyway, to put his own streamliner in business. But he still wanted diesel power.

We would have preferred to spend another year or two taking the "bugs" out of Mr. Kettering's engine, but Mr. Budd's insistence finally won us over. In June 1933 we agreed to build an eight cylinder, 600-horsepower diesel engine for his Pioneer Zephyr. When it was put in test operation in April 1934 it broke down continually, as we had feared. However, the defects were gradually ironed out of it, and in June 1934 Mr. Budd ordered two more 201A General Motors Diesels, as they were called, for his Twin Zephyrs. Meanwhile, the Union Pacific had not waited for the delivery of its streamliner. Before this, it had placed a new order with Winton in late June 1933, this time for a twelve-cylinder, 900-horsepower diesel for a six-car articulated sleeping-car train; and again in February 1934 the Union Pacific ordered six 1200-horsepower diesel passenger units for its "City" series.

These early diesel-powered streamliners were spectacular successes. In a memorable test run from Denver to Chicago, the Burlington Zephyr averaged 78 miles per hour for a total running time of only thirteen hours and ten minutes. The Union Pacific "City" trains cut the running time from the West Coast to Chicago from over sixty to less than forty hours. Operating costs to the railroads were lower and passenger patronage was considerably higher. Both of our customers immediately began calling upon us for more power so they could lengthen their trains. In May 1935 we began delivering the Union Pacific's 1200-horsepower diesels; we furnished the Burlington with two engines of 1200-horsepower apiece. These engines were able to pull twelve-car trains.

One day early in 1934 Mr. Kettering and Mr. Hamilton paid me a visit, and we got to talking about the diesel. Mr. Hamilton, who was always in close touch with the railroad people, told me that our engines were considered by them to be a vast success. However, he said, the railroads were beginning to ask General Motors to supply them with all-purpose diesel-powered locomotives instead of merely engines for power cars. Mr. Kettering indicated that he would like to undertake the development of an experimental diesel-powered locomotive. I inquired how much money he thought he would need. Mr. Kettering said that he thought it might take as much as $500,000. I told him that my own experience with new development projects suggested strongly that he could not give us a new locomotive on such a comparatively modest sum. "I know," he replied amiably, "but I figure if we spend that much, you'll come through with the rest." He got the money.

Actually, we were a long way from being in the locomotive business at that time. Our only production facilities were those for making engines in the Winton plant, and even these were somewhat outmoded; we had nothing at all for building electrical transmission equipment and locomotive bodies. Accordingly, we decided early in 1935 to build our own factory at La Grange, Illinois. This plant originally produced only the body of the locomotive—the cab and the truck—with the engines coming from Winton and the other components from outside suppliers, as before. But the La Grange plant was designed so that we could expand its operations to produce and assemble all the parts of a locomotive. We began this expansion soon after the plant was completed. By 1938 La Grange was a fully integrated locomotive plant.

Our early experience with the diesel was, as I have indicated, in the passenger-locomotive field. But in the mid-thirties Mr. Hamilton and his group decided that there was a great economic potential for diesel-powered switching locomotives. At that time one of our competitors was offering the railroads a diesel-powered switcher that weighed about one hundred tons and sold as high as $80,000. The locomotive was, in large measure, built to the customer's specifications. It was Mr. Hamilton's contention that if the customer was willing to accept a standard diesel switcher "right off the shelf," then we could market one for $72,000. Under his prodding we began to build these switchers. Indeed, we put fifty of them in production before we had one firm order.

The importance we attached to this new policy may be gauged by a memorandum written on December 12, 1935. It was from Mr. Pratt to me, and it said, at one point:

There is one fundamental policy which we believe will have to be maintained, namely, that the Electro-Motive Corporation will build a standardized product and not undertake to build to the many different standards and specifications on which each railroad demands to purchase; and our recommendation is that the policy of building a standard product be given at least a fair trial before we yield to obtaining business by letting each railroad write its own specifications as to what the locomotive should be.

As it turned out, the issue was settled very quickly. Our first batch of switchers was sold easily, deliveries beginning in May 1936. Although the margin of profit was small at first, it was enough to make a big difference in Electro-Motive's profit picture. Mr. Hamilton promised the railroads that, as our volume in switchers increased, we would pass along our operating economies to them in the form of price reductions. By 1943 when the War Production Board took General Motors out of the switcher field and directed us to concentrate entirely on freight locomotives, we had built 768 switchers; and the price to our customers on the 600-horsepower switchers was down to $59,750 by October 1940.

Meanwhile, our passenger-locomotive business expanded rapidly. By 1940 we had about 130 diesel-powered passenger locomotives in service on railroads all over the country. We began to build freight locomotives in 1939. There was an interruption during the early part of World War II when our plant was virtually out of the locomotive business while producing LST engines for the Navy.

At this point the reader may be wondering what the rest of the locomotive industry was doing while we were pushing ahead with our diesel program. With only a few exceptions and qualifications, the answer is that the rest of the industry was sticking with steam power. Though a few attempts were made, in this country and Canada, to build diesel passenger locomotives before 1940, production never advanced beyond the prototypes. (In 1940 a diesel-powered passenger locomotive built by a competitor finally went into service.) Outside of one attempt made by a group of builders in the late twenties, no manufacturer in this country, other than ourselves, brought out a diesel-powered freight locomotive until after World War II. Aside from switchers, it might be said, we were first everywhere on the railroads of this country with diesel power. To suggest, as a Senate subcommittee did in 1955, that we shoved ourselves into the locomotive market by main force, is to ignore the fact that other manufacturers failed to see the potential of the diesel. As Mr. Kettering once remarked during another congressional investigation, our biggest advantage in the locomotive industry was the fact that our competitors thought we were crazy.

Yet the superiority of diesel power over steam was apparent from the beginning. Rudolph Diesel first mentioned this superiority in railroad applications in 1894 and numerous times afterward. During the late 1920s engineering and railroad journals were carrying full reports and operating-cost data on diesel locomotives then in operation in Europe. To anyone who would listen, we could prove that the diesel offered smoother, faster, cleaner service, and an enormous saving in fuel and other operating costs. The railroads, which were eager to trim their operating costs in every way possible during the 1930s, listened eagerly; the other locomotive manufacturers continued to regard the diesel as a sort of passing fad. This explains why a group of long-established, economically strong locomotive manufacturers, with strong ties to their customers, were so easily outdistanced by one newcomer to the business. It was not until the mid-1950s that the building of steam locomotives in this country stopped completely, with production in the closing years going largely to export. Less than a hundred steam locomotives remain in operation in the United States today. Diesel power alone is now being purchased by the railroads, except for electric locomotives used on electric-powered roads. This revolution in the railroad industry in the United States was made very largely by General Motors.

It is hard to make precise statements about the future of the diesel locomotive business, but it appears that the market in the United States will be somewhat smaller in the years ahead. Railroad passenger service is being discontinued in many areas of the country, and even freightcar loadings have declined somewhat in recent years. There were about 60 per cent more steam locomotives in service during the mid1930s than there are diesels today. This fact reflects the greater power and operating availability of the diesel, of course, but it also reflects the depressed condition of the railroads.

Overseas there still are some 100,000 steam locomotives in operation. These eventually will be replaced by diesel-electric, diesel hydraulic, and electric locomotives. The potential market for diesel electric locomotives overseas is approximately 40,000 units. The Electro-Motive Division has developed a wide range of lightweight, restricted-clearance locomotives to meet this export demand. Where applicable, standard domestic locomotives have been sold overseas. Over four thousand General Motors locomotives are now in service in thirty-seven countries outside the United States—nine countries, including Canada, in the Western Hemisphere and twenty-eight countries of the Eastern Hemisphere.

The U.S. market is now a replacement, reconditioning, and upgrading rather than a new-user market. The so-called upgrading market is, of course, an increasingly important one today, and I do not mean to minimize it. Still, the industry in the United States has been dieselized; the revolution is over. At the same time, it is just under way overseas.

Frigidaire

Despite a lack of enthusiasm at the highest levels of the corporation in the early days, the Frigidaire Division has grown steadily for about forty-five years and has become a major factor in the appliance industry. The Frigidaire line today includes electric household refrigerators, food freezers, ice-cube makers, automatic clothes washers and dryers, electric ranges, water heaters, dishwashers, food-waste disposers, air-conditioning equipment, and commercial laundry and dry-cleaning equipment. Frigidaire now has about ten thousand outlets in the United States.

The curious story of how General Motors got into the refrigerator business begins in June 1918 when Mr. Durant, who was then president of the corporation, purchased the Guardian Frigerator Company of Detroit. Mr. Durant made the purchase in his own name and with his own funds; the precise amount was $56,366.50. The company passed from Mr. Durant to General Motors in May 1919 at the same price. It was a small enterprise of no great substance. He soon renamed the company the Frigidaire Corporation, and also gave the name Frigidaire to the rather crude, primitive device which was then its sole product. Mr. Durant's motives in this transaction are not within my knowledge. But he was, of course, a man of boundless enthusiasms and great curiosity; and it is easy to understand that an "iceless frigerator"—as the Guardian product was called—would excite both of these qualities. I can only admire his gift for being in touch with future developments in this as well as the automotive field.

While I had no personal knowledge of Mr. Durant's transaction at the time it took place, John L. Pratt has told me that in his opinion more than enthusiasm for a new appliance underlay the purchase. He says that Mr. Durant was concerned about the prospect of the automobile business being declared unessential to our World War I mobilization effort, and was looking for an "essential" business to take the place of civilian automobiles. Given the great national effort to conserve food during World War I, a refrigerator company might be considered essential. However, the government made no effort to end automobile production; and in November, five months after his purchase had been made, the war ended.

The original Guardian refrigerator had been built by a Dayton mechanical engineer named Alfred Mellowes in 1915. The following year he organized the Guardian Frigerator Company in Detroit to manufacture and sell his device. Between April 1, 1916, and February 28, 1918, Guardian built and sold only thirty-four refrigerators, all of which were installed in homes in the Detroit area. Guardian's manufacturing facilities in 1917 consisted of only two lathes, one drill press, one shaper, one power saw, and a hand vacuum pump. In addition to manufacturing the "frigerators," Mr. Mellowes personally serviced them; he kept in close touch with the purchasers, visiting each of them every two or three weeks. As we ascertained at the time we bought Frigidaire, most of these early Guardian customers were pleased with the product. Many of them had, in fact, despite the numerous service problems, invested in Mr. Mellowes' company. But as investors, it appeared, they were less happily situated than they were as consumers. During its first twenty-three months Guardian showed a loss of $19,582. In the three months just before Mr. Durant bought it the company lost another $14,580, bringing its total deficit to $34,162. Less than forty refrigerators had been built and sold in the entire period. It is not difficult to understand why the original shareholders were happy to sell out.

When Frigidaire passed into General Motors, we tooled up in our Northway plant in Detroit to manufacture Frigidaire Model A— a machine which was identical to the old Guardian except for minor mechanical changes. Our miscalculation about the product's suitability for mass consumption was speedily brought home to us. Model A, and its successors in the first few years, remained a luxury product. What was worse, we could not get the "bugs" out of the machine, which broke down repeatedly. Our efforts to introduce a sales and service organization into a number of cities outside of Detroit were largely unsuccessful. It appeared that the machine really needed the kind of steady personal service that Mr. Mellowes had provided his small group of customers; but this kind of service was obviously impossible in a product intended for a mass market. After about a year and a half we seriously considered whether the Frigidaire operation might not be jettisoned. Something of our frame of mind may be sensed from the minutes of a meeting which took place in my office on February 9, 1921. The summary of my remarks includes these comments:

Frigidaire Corporation: Located at Detroit, Mich, and makes Frigidaires which up to the present have been a failure. Models have been changed frequently in order to create demand, but without success. Branches were opened at various points which have since been discontinued . . . Loss to date about $1,520,000. Inventory is about $1,100,000 —total loss expected to run about $2,500,000.

In a year when General Motors was in serious need of operating capital, the continued losses and relatively high inventory could not long be tolerated. And it is possible that Frigidaire would somehow have been disposed of then except for one fortuitous circumstance, upon which hangs a story.

In an earlier chapter I told how General Motors in 1919 acquired the Dayton properties with which Mr. Kettering was associated. Among these properties were the Domestic Engineering Company and the Dayton Metal Products Company.

The Domestic Engineering Company—later renamed the DelcoLight Company—was a manufacturer of home-lighting plants, which were sold mostly to farmers.

The Dayton Metal Products Company, an armament manufacturing concern, had begun research in the refrigeration field early in 1918 as part of a program designed to obtain a product which might keep the company in operation when the war ended and the armament business ceased.

The two enterprises—Domestic Engineering and Dayton Metal Products—were in the appliance business in some items, and were preparing to expand into some other items. With these enterprises General Motors also acquired all of the refrigeration developments of Mr. Kettering's research group. This informal research organization continued operations at Dayton until June 12, 1920, when the subsidiary General Motors Research Corporation was organized. General Motors thus acquired some outstanding engineers in this field, as well as the management and sales ability of Richard H. Grant, who was to contribute importantly to the success of Frigidaire in the early and middle 1920s.

All of these factors came together in our decision during the slump of 1921 to continue with Frigidaire. It was clear that we had at Dayton the research background and an organization to back up the Frigidaire development. Delco-Light had available a fine sales force spread over large areas of the country, and some unused manufacturing capacity which could be made suitable for the production of refrigerators. So we moved Frigidaire to Dayton, combined its operations with those of Delco-Light, and started on a new course in the refrigerator industry on a larger scale than thereto- fore.

The decision proved to be a sound one. Frigidaire's heavy losses in 1921 were reduced steadily in the next two years, and in 1924 the operation showed a profit for the first time. Meanwhile, production rose rapidly. Only a few more than a thousand units had been produced in 1921 at the Northway plant; about 2100 were sold in 1922, the first full year of operations at Dayton. The figure rose to 4700 in 1923, 20,200 in 1924, and 63,500 in 1925. By the last year, Frigidaire was established as a leading factor in the new refrigerator industry; it represented, I believe, more than half of the market. By 1927 it was apparent that Frigidaire was becoming much too big to be operated within Delco-Light, and in January 1928 it was removed from that company. Part of its operations had already been moved to nearby Moraine, Ohio, where we had a plant available. Frigidaire became a division of General Motors in December 1933.

Once we had decided to build up Frigidaire we made a number of major ground-breaking advances in the design and manufacture of the machine. Without these contributions, it is safe to say, popular acceptance of the refrigerator would have been delayed for a considerable period of time.

As I have indicated, the Guardian organization originally had no real research staff outside of Mr. Mellowes himself. Even in 1921, when Frigidaire was moved into Delco-Light, there were only twenty-odd engineers, model makers, testers, and the like engaged in this work. We realized that the whole future of Frigidaire depended on our ability to crack several research problems, and to produce a machine that would operate safely, economically, and dependably; hence we placed great emphasis on research. We soon managed to get rid of the space-consuming brine tank and water-cooled compressor used on the original Guardian machine; these devices, which were major sources of refrigerator breakdowns, were replaced by a direct-expansion coil and a two-cylinder, air-cooled compressor. In the early machines, food was sometimes contaminated when moisture leaked into the refrigerator; we overcame this problem by introducing asphalt-and-cork sealing. We reduced the weight of the machine and considerably improved its appearance when we introduced the all-porcelain cabinet in 1927. All of these improvements were instrumental in the great expansion of the Frigidaire market during the 1920s. Another major cause of this expansion was our ability to get prices down. The 1922 B-9 wood refrigerator with brine tank and water-cooled compressors had a net weight of 834 pounds and sold for $714. In contrast, the M-9 Frigidaire model of 1926, a steel cabinet fitted with an air-cooled compressor and direct-cooling coils, had a net weight of 362 pounds and sold for $468.

During the 1919-26 period no other manufacturer or organization made any appreciable contribution to the refrigeration business in research, engineering development, mass-production methods, or distribution and servicing techniques. Our biggest research problem in Frigidaire, and the corporation's great ultimate contribution, concerned the refrigerant itself. The fact was, during the 1920s, that the refrigerants used by Frigidaire, and by all its leading competitors, had some health hazards; fumes from the refrigerating agents were toxic and in a few cases had actually caused the death of persons who breathed them. Because of the health hazard, these early refrigerators were sometimes kept on the back porch rather than in the kitchen; hospitals generally could not use them at all. We believed that sulphur dioxide, the agent first used in our refrigerators, was the least dangerous of the known refrigerants—principally because its distinctly irritating odor served as a warning to anyone breathing it. Nevertheless, it was clear that, ultimately, something better had to be found.

In 1928 Mr. Kettering, who was then director of General Motors Research Laboratories, initiated a major assault on the whole problem of the refrigerating agent. He commissioned one of his former associates in General Motors, Thomas Midgley, Jr.—the man who had developed tetraethyl lead—to find a new agent. After a series of conferences between Mr. Midgley, Mr. Kettering, and Frigidaire executives, they agreed that the refrigerant they were looking for should meet certain requirements. These were:

Of primary importance:

(1) To have a suitable boiling point.

(2) To be nonpoisonous.

(3) To be nonflammable.

(4) To have a distinct but not unpleasant odor.

Of secondary importance:

(5) To be immiscible with lubricating oils.

(6) To be relatively inexpensive.

These "secondary" requirements, it was understood, would be met so long as they did not conflict with the primary requirements. But there was agreement that all of the first four specifications had to be met before the electric refrigerator could be regarded as a complete success. A study of all existing literature was made at the Research Laboratories, under Mr. Kettering's direction, for compounds winch might meet these specifications. This study pointed out the possibility of using fluorinated hydrocarbons. All through 1928 Mr. Midgley and some associates, especially Dr. A. L. Henne, worked in a private laboratory in Dayton in an effort to find a suitable refrigerant. They soon came to believe that some of the chlorofluoro derivatives of methane might do the job. By the end of the year Mr. Midgley had determined that dichloro-difluoro-methane, called Freon-12, would meet all four of the primary requirements agreed upon. It would not meet either of the two secondary requirements, but since it was clearly the best refrigerant available, Mr. Midgley and his associates began working on the development of processes for manufacturing the compound. A pilot plant was designed and put in operation at Dayton during the fall and winter of 1929-30.

In the fall of 1929 we knew as much about the Freon-12 refrigerant as we had to know. Frigidaire chemists had made exhaustive studies of the compound's physical properties. They had determined the corrosion effects of Freon-12 on high and low-carbon steels, aluminum, copper, monel metal, tin, zinc, tin-lead solders, and other metals and alloys used in refrigerating systems. They had examined the effect of Freon-12 on different foods, and on flowers and furs. The tests were satisfactory to us. At the 1930 meeting of the American Chemical Society, Mr. Midgley read a paper on Freon-12 and publicly demonstrated that it was nonflammable; he proved that it was nontoxic by inhaling some of it himself.

As I have indicated, Freon-12 did not meet either of Mr. Midgley's two secondary requirements. It was quite expensive, in fact. Whereas sulphur dioxide had cost six cents a pound, the initial price of Freon-12 was sixty-one cents in 1931. Even now it costs more than sulphur dioxide did then—but health-department codes do not allow the use of the latter.

Since we regarded our new compound as the safest refrigerant available, we offered it to our competitors from the beginning, and by the mid-1930s Freon-12 was used almost universally in electric refrigerators. Even today, no better refrigerant has been found.

By 1932 or thereabouts it was unmistakably clear to us that in Frigidaire we had a property of vast growth potential. In 1929 we had manufactured our one-millionth Frigidaire, and three years later we had manufactured 2,250,000. Our success in developing Freon-12 removed the last roadblock standing in the way of the refrigerator industry. But while it was clear that Frigidaire and the industry would expand, it was also clear that Frigidaire's share of this great market must inevitably decline somewhat. Several companies would begin making refrigerators toward the end of the 1920s. Kelvinator was, of course, a pioneer. The original Kelvinator Corporation entered the electric-refrigerator field in 1914 and was the first enterprise to manufacture mechanical refrigerators for household use on a commercial scale. General Electric and Norge entered the field in 1927, Westinghouse in 1930. By 1940, the last prewar year of unregulated commercial production, Frigidaire's share of the refrigerator market—which had been above 50 per cent in the 1920s—was down to 20 to 25 per cent. But our smaller percentage represented a larger volume. Shipment of our refrigerators rose from some 300,000 in 1929 to 620,000 in 1940.

During the years 1926-36 a number of Frigidaire's competitors gained an advantage over us in the marketing area. They began to make and sell radios, electric ranges, washers, ironers, and dishwashers, while Frigidaire concentrated on refrigerators. In 1937 we added electric kitchen ranges to the Frigidaire line, and a few years later, window-type room air-conditioners. But these did little to overcome Frigidaire's competitive disadvantage. Obviously, families and home builders who wanted to purchase a full complement of household appliances would buy from one of the manufacturers who offered a complete line.

We failed to expand the Frigidaire line in the years before the war. As early as 1935, for example, Mr. Pratt had suggested that Frigidaire get more actively into air-conditioning; but his suggestion did not register on us, and the proposal was not then adopted.

During the war we made a review of Frigidaire's prospects and concluded that it would no longer be feasible to operate in the appliance field on a limited basis. A survey of Frigidaire dealers conducted prior to the end of the war served to fortify this conviction. In response to the survey question, "Should Frigidaire manufacture additional appliance products?" 99 per cent of the dealers who were polled replied, "Yes." The dealers indicated that, principally, they wanted automatic washing machines, refrigerator-freezer combinations, conventional washing machines, food freezers, gas ranges, and ironing machines—in that order.

Most of these appliances and several others were added by Frigidaire in the postwar years. The following list shows the years in which we introduced new household appliances:

Home food freezers	1947
Automatic washers	1947
Dryers	1947
Automatic ice-cube makers	1950
Dishwashers	1955
Wall ovens	1955
Fold-back cooking units	1955
Built-in cooking units	1956

Meanwhile, our original product—the refrigerator—has been enlarged and improved little by little to such an extent that it has become almost a new appliance. The typical refrigerator sold in the early 1930s was a five-cubic-foot model, styled rather drearily, and depressingly bulky in relation to its actual refrigeration space. Refrigerators sold today have, as a rule, from ten to nineteen cubic feet of storage space. They are beautifully styled, require no defrosting, and have considerable freezer space. There is no question that the modern refrigerator is a much better buy than its early counterpart. I am indebted to a study by Professor M. L. Burstein of Northwestern University for some detailed data bearing on this point. He has calculated that "the real price of refrigeration services in 1955 was but 23 per cent of that in 1931." That comes pretty close to the essential meaning of progress.

Aviation

General Motors has been involved in the aviation industry in several different ways. The bulk of our aviation business has been military, of course, and has consisted of work done under contract for the federal government—most of it during World War II and in the ensuing years of the cold war. But that is not the whole story.

It will, I suspect, come as a surprise to many readers that General Motors long ago made a major effort to enter the commercial aviation field. Bendix Corporation, North American Aviation, Trans World Airlines, and Eastern Air Lines, all owe something of their present identities to the activities of General Motors.

Our venture into commercial aviation was made in 1929. In that year we made two large investments and one small one in aviation. We purchased a 24 per cent interest in the newly formed Bendix Aviation Corporation and a 40 per cent interest in the Fokker Aircraft Corporation of America. Together, these investments cost us some $23 million. In addition, we purchased the entire capital stock of the Allison Engineering Company. This investment cost only $592,000, and did not play an important part in the plans we had then to enter the aviation industry.

Our 1929 decision to get into aviation has an interesting background. I should mention that General Motors was not entirely a stranger to the aviation industry at that time. During World War I, Buick and Cadillac had combined to manufacture the famous Liberty aircraft engine for the government, along with Ford, Packard, Lincoln, and Marmon. More than 2500 of these engines were actually produced by us, and orders for over 10,000 more were on our books at the time of the 1918 armistice. From an engineering standpoint, there was not a great deal of difference in those days between an airplane engine and an automobile engine, and we were able, in consequence, to make good use of our automotive experience in compiling an outstanding production record. In addition, General Motors acquired in 1919 the Dayton Wright Airplane Company, which had produced a total of 3300 airplanes during the war period. Fisher Body also—before its purchase by General Motors—was an important manufacturer of military airplanes.

During the 1920s it became steadily clearer that aviation was to be one of the great American growth industries; and especially after Lindbergh's dramatic flight in 1927 there was a vast public enthusiasm for aviation and a widespread conviction, which we shared, that it would soon accomplish many more "miracles." As automobile producers, we were especially concerned about one possible use of the airplane. There was in the late 1920s a great deal of talk about developing a "flivver" plane—that is, a small plane for everyday family use. We knew, of course, that any such plane would have to be much safer than existing models and also much cheaper. But as one aviation miracle succeeded another, our conviction grew that the flivver plane was at least a possibility. The development of such a plane would have large, unforeseeable consequences for the automobile industry, and we felt that we had to gain some protection by "declaring ourselves in" the aviation industry. In 1929 we did not plan to operate either Bendix or Fokker as a division of General Motors; our investments were made as a means of maintaining a direct and continuing contact with developments in aviation. Our 1929 annual report to shareholders summed up our thinking on the matter as follows:

. . . General Motors, in forming this association [with the aviation industry], felt that, in view of the more or less close relationship in an engineering way between the airplane and the motor car, its operating organization, technical and otherwise, should be placed in a position where it would have an opportunity to [come into] contact with the specific problems involved in transportation by air. What the future of the airplane may be no one can positively state at this time. Through this association General Motors will be able to evaluate the development of the industry and determine its future policies with a more definite knowledge of the facts.

As these words suggest, the engineering techniques of the automobile and aircraft industries were still quite similar in 1929—much more similar than they are today. Thus in acquiring our interests in the aviation companies we also gained access to some valuable technical information that was directly relevant to our own automobile operations. Bendix, especially, owned or controlled some important patents for devices applicable to the automobile industry. Indeed, its accessory lines included some automobile components— for example, brakes, carburetors, and starting drives for engines. The company had a superb technical staff—a fact winch made our investment all the more attractive. Our principal contributions to both Bendix and Fokker, after we had made our investment in these companies, were in the realm of corporate organization and management.

Our 40 per cent interest in Fokker cost us $7,782,000. This company had two small, leased plants at the time we made our investment: one in Hasbrouck Heights, New Jersey, and one in Glendale, West Virginia. Anthony H. G. Fokker, a brilliant Dutch aircraft builder, had formed the company some time earlier to exploit the American manufacturing rights to his work. His aircraft had figured prominently in the pioneer days of aviation; they had been involved in such historic events as the first nonstop flight across the United States, Byrd's flight over the North Pole, and the first flight from the United States to Hawaii. When we bought into Fokker, the company was engaged primarily in aircraft manufacture for the United States government and, to a lesser extent, for commercial air-transport operators. Soon after our investment, the company incurred some serious operating losses. We felt that these losses reflected weakness in the company's management, and we conveyed our views to Mr. Fokker. He did not agree with us, but after a series of exchanges he withdrew from the company and returned to Holland. We then embarked upon a course of action which completely transformed the character of the organization.

The following interrelationships are complex, and I see no way to simplify the description of them. First of all, we changed the name of the Fokker Aircraft Corporation of America to General Aviation Manufacturing Corporation and consolidated the operations in a leased plant in Dundalk, Maryland. In April 1933 we took another important step. We combined General Aviation and North American Aviation; all of General Aviation's assets were exchanged for approximately 1,500,000 shares of North American common stock. General Aviation was subsequently liquidated, and its holdings in North American stock distributed to the shareholders. As a result of this distribution, and of open-market purchases for our own account, General Motors' equity in North American amounted to nearly 30 per cent of that company's outstanding stock by the end of 1933.

North American Aviation had been organized as a holding company in 1928. Although it had some substantial investments in aircraft manufacturing companies even before it was joined with General Aviation, its primary concern had been the airline business. It owned all of Eastern Air Transport (later called Eastern Air Lines) , 26.7 per cent of Transcontinental Air Transport, and 5.3 per cent of Western Air Express Corporation. General Aviation had also owned 36.6 per cent of Western Air Express stock. Afterward, therefore, North American owned 41.9 per cent of Western Air Express stock. Furthermore, Western Air Express and Transcontinental Air Transport each owned 47.5 per cent of the stock of Transcontinental and Western Air, Inc. (now Trans World Airlines). The upshot of the arrangement, then, was that General Motors held a 30 per cent interest in North American, and North American held about a 33 per cent interest in TWA. North American was thereby enabled to co-ordinate the transcontinental operations of TWA with the East Coast system of its own Eastern Air Lines.

The Air Mail Act of 1934 prohibited the ownership of stock in an airline company by companies engaged either directly or through a subsidiary in aircraft manufacturing. North American therefore distributed its holdings in TWA to its shareholders. As a shareholder of North American, General Motors received some 13 per cent of TWA's stock, which we sold in 1935.

For a time North American operated Eastern Air Lines as a division, and then disposed of this airline in March 1938. As the largest single shareholder in North American, General Motors had several representatives on its board of directors. One day, during the period when North American was negotiating the sale of Eastern Air Lines to some Wall Street interests, I received a telephone call from Eddie Rickenbacker, the United States' great World War I flying ace. He had been active in the Eastern Air management and was now interested in bidding to buy control of the airline. He complained that he was not being given a chance, however, and asked if I would intervene in his behalf.

I had always considered Eddie to be a capable operator, and I naturally wanted him to have an equal opportunity to bid for Eastern; I felt that he could be counted on to develop an efficient operation. I told him I would see what I could do. The next morning I looked into the matter and found that the Eastern Air stock had not yet been sold. I made a plea in Eddie's behalf, and as a result he was given thirty days to get the backing which would enable him to bid.

He did not get his backing very easily, however, and as the deadline approached he became understandably nervous about the outcome. The next to the last day fell on a Saturday. Eddie called me at my apartment as I was preparing for bed, and inquired if he could come over for a few minutes. When he arrived he indicated that his prospects for getting the money were still excellent, but that he might need more time. He wondered if he could have a few days' extension. I told him not to worry, and he left in good spirits. But as it turned out, he did not need the extension. His backers called him the next morning and told him that they were prepared to go through with the deal. This disposal by North American of its Eastern Air Lines operation was a transaction that gave us all a great deal of satisfaction.

In the reorganization that followed the Air Mail Act of 1934, North American became an operating company. Its manufacturing operations were consolidated and moved to a new plant in Inglewood, California. During the ensuing years, the company placed emphasis on the development of military aircraft and made some notable strides in that direction. In the late 1930s the company won several military design competitions, and these successes established it as one of the nation's leading aircraft manufacturers.

A number of aircraft which evolved from this early development work played a vital part in World War II. Among the more famous of the North American planes were the P-51 Mustang fighter—perhaps the most highly regarded fighter plane employed by the Allied forces during the war; the B-25 Mitchell bomber used by General Doolittle in his historic raid on Tokyo; and the ubiquitous AT-6 Texan trainer, which became virtually standard equipment at Air Corps and Navy training bases and was used extensively by other Allied countries.

The AT-6, by the way, reflected the General Motors influence on North American. As automobile men, we naturally thought in terms of "standardized" production models which could realize the inherent economies of volume production. North American began looking for a plane that could be marketed this way and soon decided that a good basic-training plane was the best bet. Even before the war the AT-6 became its "bread and butter" model.

General Motors was continuously represented on North American's board of directors from 1933 until we finally disposed of our interest in 1948. During that time—and especially in the earlier years—we provided a considerable amount of policy and administrative guidance through our representatives on the board, and we were instrumental, I believe, in developing an efficient, systematic approach to management in the company. North American's corporate organization and its financial, production, and cost controls were our special contributions. It appears that in 1939 North American was the only aircraft-manufacturing company with production and cost-control systems like those used in the automobile industry.

A major share of the credit for introducing General Motors' management techniques at North American, and at Bendix, too, must go to Ernest R. Breech. Mr. Breech was originally a financial man in General Motors (he was general assistant treasurer from 1929 to 1933), but when he moved over to North American he soon showed a great talent for operations as well. He was chairman of the board of North American Aviation from 1933 to 1942—the years during which the holding company was converted into a large manufacturing operation. In addition, he became a director of Bendix in 1937. I had always considered Mr. Breech to be an excellent prospect for top management, and had tried for some time to bring him into a good operating position in General Motors. I was opposed in this effort by William S. Knudsen, executive vice president and later president of General Motors, who still regarded Mr. Breech as a financial man. But finally, in 1937, I found a spot for him as group executive in charge of General Motors' household appliance operations. He filled this post with distinction while continuing to serve as chairman of North American and as a Bendix director.

In 1942 he became president of Bendix, relinquishing his other assignments. At Bendix again, he performed brilliantly all during the war years and more than justified my faith in him. But the story of his career has an ironic twist, as many persons know. In performing so well in all his General Motors assignments, he attracted the attention of Henry Ford II, who wanted someone to head the rebuilding program of the Ford Motor Company. Mr. Breech got that job in 1946, and introduced General Motors' management and financial techniques into the new modern Ford organization.

When Mr. Breech was chairman of North American, he induced J. H. ("Dutch") Kindelberger, who had been chief engineer for Douglas Aircraft, to head up operations. Mr. Kindelberger was elected president and chief executive officer of North American at the end of 1934. He was an extremely capable engineer and demonstrated great technical competence in aircraft design and manufacture. He developed into a fine administrator, and came to be recognized as a man who could produce outstanding military planes at low cost. But he had had very little general administrative experience before coming to North American, and, recognizing his own limitations, he relied at first on the General Motors directors for advice and counsel. Messrs. Breech and Kindelberger, together with Henry M. Hogan, then assistant general counsel of General Motors, constituted a sort of informal executive committee and regularly consulted each other on all the important company problems which arose between meetings of the board of directors. Messrs. Breech and Hogan in turn reported to Albert Bradley or C. E. Wilson, who, in addition to their regular duties as executive officers in General Motors, had group responsibility for our investments in associated companies.

Our relations with Bendix were much the same as with North American. We were represented on the Bendix board of directors from 1929 until 1937 by Messrs. Wilson and Bradley; the latter was also chairman of the Bendix Finance Committee throughout that period. In 1937 the press of other duties forced these two to give up their Bendix directorships, and they were succeeded on the board by Mr. Breech and by A. C. Anderson, the comptroller of General Motors. Our representatives on the Bendix board took a direct interest in the internal management of Bendix and were instrumental, I believe, in improving management effectiveness. They were responsible for some organizational changes and for a new and effective system of co-ordinating the semi-autonomous divisions. Our representatives also had a direct hand in the elevation of Malcolm Ferguson to the important post of general manager of the South Bend automotive parts plant of that company. He later became president of Bendix.

By the end of the 1930s our perspective on North American and Bendix had changed considerably. Our original motive for investing in the aviation industry—the feeling that the industry might somehow produce a flivver plane which could compete with the automobile—came to seem less relevant as the years passed. No plane suitable for "family use" was ever developed; indeed, the whole commercial-aviation field remained a small one during the years of the depression. North American and Bendix continued to grow, but both companies discovered that their greatest opportunities lay in the military field. By 1940 each company had annual sales running around $40 million, and the great bulk of this was defense work done under government contracts. In 1944 at the peak of wartime production, North American's sales were about $700 million, and Bendix's sales came to more than $800 million. These huge figures suggest the far-reaching consequences for us of our original concern about the flivver plane.

The Allison Engineering Company, which we also acquired in 1929, has had a growth history no less spectacular than North American and Bendix. As I have mentioned, we purchased Allison outright for only $592,000. By our standards, it was a small operation: the company had fewer than 200 employees in 1929, and its manufacturing facilities occupied only about 50,000 square feet of floor space. We considered it to be of only minor importance in our plans to enter the aviation industry. Yet as events turned out, we were to make Allison our principal link to the industry.

When we acquired Allison in 1929 the company had been in existence for fourteen years. In its early years it was not in the aviation field; it was primarily a supporting machine shop for racing cars at the Indianapolis Speedway. Its founder, James A. Allison, gradually shaped an organization of skilled mechanics, machinists, and engineers, and began to produce a few marine engines and reduction gears for boats and aircraft. In the early 1920s Allison was able to accept a contract for the modification of World War I Liberty aircraft engines. Chronic failures in the crankshaft and connecting-rod bearings had seriously limited the durability of these engines. But Allison was able to develop a steel-backed, lead-bronze, crankshaft main bearing that was capable of supporting higher horsepower loads without failure. The company also developed an ingenious method of casting lead-bronze on both the inner and outer surfaces of a steel shell, which could be used in making connecting-rod bearings of great durability. These developments were the basis for the highly regarded Allison bearing that came to be used extensively in high-horsepower engines throughout the world. The production of this bearing, and the contracts to modify Liberty engines, accounted for the principal business of the company during the 1920s.

When Mr. Allison died in 1928 the company was put up for sale, on the condition that operations must continue in Indianapolis. Several prospective buyers were approached, but none was willing to accept this stipulation. Fortunately, C. E. Wilson had become well acquainted with the Allison organization while he was general manager of the Delco-Remy Division in Anderson, Indiana. He knew that the organization possessed valuable mechanical skills that we could use. We had no objection to continuing operations in Indianapolis, and on Mr. Wilson's recommendation we approved the purchase in early 1929. Norman H. Gilman, who had been president and chief engineer under Mr. Allison, continued in charge as general manager after our acquisition.

Early in the 1930s Allison embarked upon a project which proved to be of great military significance. This was the V-1710 engine project, and it was initiated by Mr. Gilman. After a careful survey of all the military aircraft engines then in existence, Mr. Gilman concluded that the armed services would one day require a reciprocating engine of 1000 horsepower; he also concluded that the engine should be liquid cooled (which would give it a slimmer shape than an air-cooled engine).

Only very meager funds for such projects were available from the military in the early 1930s, but Mr. Gilman did win a small contract, and Allison set to work designing the engine. A partial success was achieved in 1935 with a 1000-horsepower engine that functioned well for about fifty hours. However, the engineers could not get the engine to function for 150 hours—which army specifications required. To speed up development work on the engine we assigned Ronald M. Hazen, an outstanding engineer with the General Motors Research Laboratories, to work at Allison. Mr. Hazen's efforts were successful, and on April 23, 1937, the V-1710 passed all the tests required by the Army Air Corps. It was the first airplane engine in the United States to qualify at 1000 horsepower and the nation's first really successful high-temperature, liquid-cooled engine.

Until the V-1710 engine was developed the Air Corps had taken for granted the superiority of the air-cooled engine. But the Allison engine quickly proved its worth: In March 1939 a Curtiss P-40, powered by a V-1710 engine, won the Air Corps fighter aircraft race with a clear speed advantage of forty miles per hour over the previous winner. There was, naturally, a sudden great surge of interest in the Allison engine after that event. Not only the U. S. Air Corps, but the armed forces of Britain and France began to look closely at our product.

Allison now had a serious problem. Though we had built it up somewhat since 1929, the year of our acquisition, it was still essentially a small engineering firm, geared mainly to experimental work with no facilities at all for quantity production. And quantity production at the end of the 1930s was being desperately demanded by the government.

The Assistant Secretary of War, Louis Johnson, personally visited Mr. Knudsen, who was then president of General Motors, to see what could be done about producing Allison engines. At that time there were firm orders for only 836 engines; and, as Mr. Johnson conceded, he was in no position to assure us that more orders would be forthcoming. Viewed simply as a business proposition, building a factory to make 836 engines seemed to be a bad risk; indeed, there was the risk that some new turn in the international situation or some new technological breakthrough might wipe out even tins small demand before our factory was built. Nevertheless, we decided, after weighing the matter closely, to establish a new Allison plant in Indianapolis. This decision was based on a feeling that the V-1710 engine would probably be in great demand. Moreover, one does not lightly turn down any government request having to do with national security.

And so, on May 30, 1939, we broke ground in the shadow of the Indianapolis Speedway for a new plant to build the Allison engine. As it happened, more orders for the V-1710 did follow: the French government ordered 700 of the engines in February 1940, and the British ordered another 3500 a few months later. By December 1941 Allison was producing engines at the rate of 1100 a month. During the war we forced this rate still higher—even though the engine was being continually redesigned and repowered until it finally attained a combat rating of some 2250 horsepower. By December 1947, when we stopped making the V-1710 engine, total production was up to 70,000 engines. They had performed brilliantly all during the war and were used on such famous fighter planes as the Curtiss P-40 Warhawk, the Bell P-39 Aircobra, the Bell P-63 King Cobra, and the Lockheed P-38 Lightning.

Early in the war it became apparent that our involvement in aviation was so large as to raise a question about our permanent place in the industry. Accordingly, we made an effort to redefine our thinking about aviation and the part we should play in it. The principal statement on this important matter is a report which I made in 1942 to the Postwar Planning Group in General Motors. The recommendations in this report were eventually adopted by the corporation's Policy Committee, and they became the basis for our postwar aviation program.

In the report I indicated that there would be three major markets in the postwar aircraft industry—military, commercial air transport, and private civilian flying. I then raised the question whether we wanted to participate in any or all of these markets as producers of complete aircraft. I pointed out that the manufacture of military aircraft would involve a large amount of engineering and development work with continuous modification of low-volume models. Moreover, there would undoubtedly be excess capacity in the industry, resulting in severe competition for whatever business did exist, with little prospect for anything more than a small margin of profit.

In the commercial transport field I foresaw a rapid acceleration of transportation by air, not only for passengers but for freight as well. Even in this expanded market, however, the potential sales volume available to a manufacturer would be limited. I assumed that there would be something on the order of a tenfold increase in the number of transport planes in use—roughly 4000 planes altogether. But with an average life span of perhaps five years for each plane, the potential volume available to a single producer in any one year would not be great.

I was also dubious about the advisability of our manufacturing small private planes. While I believed that there would be some postwar expansion in the market for these planes—for both business and personal use— I felt that growth in this area would be limited until technology had advanced to a point where a far greater degree of safety could be attained. I stated that, unless there was some revolutionary breakthrough on safety, the private plane would not become a serious competitor of the motorcar in the foreseeable future.

In short, none of the three aircraft markets seemed inviting to General Motors. Furthermore, I said, if General Motors got into the business of manufacturing complete aircraft, we might jeopardize the other aviation business of the corporation. Our Allison Division was, and would continue to be, a major producer of airplane engines and certain aircraft accessories. In general, these accessories, with comparatively minor variations of an engineering and production character, would be applicable to many kinds of planes and might normally account for some 40 to 45 per cent of the cost of a complete airplane. The sales potential in this area of the market was substantial. But in order to realize this potential, an accessory manufacturer would need the engineering co-operation and confidence of the aircraft manufacturers who were his customers. If we were producing complete aircraft ourselves, we would have difficulty establishing such a relationship with our customers. How could we expect an airframe manufacturer engaged in developing a new plane to disclose its forward designs to an accessories producer who had the opportunity to make use of these designs as a competitor? In short, it seemed incomprehensible to me that we could expect to sell our accessories successfully and at the same time compete with those to whom we must sell them by producing one or more types of airframes ourselves.

Discussion of this subject continued for some time and on August 17, 1943, our postwar aviation policy was formally defined in the following resolution of the corporation's Policy Committee:

First: The corporation should not contemplate the production of complete airplanes in either the military or transport areas.

Second: The corporation should develop as complete a position in the manufacture of accessories as its capacity and circumstances make possible.

The reader will observe that we did not, at this time, specifically exclude the possibility of our manufacturing a small plane for private business and personal use. We still doubted that volume production of such a plane would be possible on a basis attractive to General Motors; however, we felt we could not ignore the possibility completely. In my report I recommended that we establish a program to keep abreast of technological developments in the small-plane business, but we later discarded the idea as impracticable. However, North American Aviation did go on to design and manufacture a plane for individual transportation, the Navion.

The formulation of our postwar aviation policy naturally had considerable bearing on our attitude toward our investments in North American and Bendix. During the war North American became one of the nation's leading airframe manufacturers, and we concluded that a continued investment in the company would be no less damaging to General Motors' accessory business than direct manufacturing of airframes ourselves. Furthermore, it became increasingly clear that General Motors could not employ its mass-production techniques effectively in the airframe industry. We decided, therefore, that it would be in the best interests of both General Motors and North American to dispose of our holdings in the company at some appropriate time.

Bendix was a somewhat different proposition. This company already held a strong position in the aviation accessories field, and its activities fitted in very nicely with our own scheme of operations and postwar policy objectives. At one point we gave very serious study to acquiring Bendix outright and operating it as a consolidated division or subsidiary of General Motors, but decided against it. We gradually arrived at a general policy of disposing of minority interests, and in 1948 sold our interests in North American and Bendix. The capital thus released was employed in our rapidly expanding automotive operations.

Our contributions to Bendix and North American during the period of our association were not in the engineering and technological fields. They were in the more intangible area of business management. To the extent that our management philosophy was imparted to these companies and to the aviation industry in general, General Motors, I think, made a tangible contribution to that industry.