Late in December 1921, having in mind the breakdown of the test at Oakland and the problems that the proposed new car had created, I wrote some notes to myself to clarify in my mind the problems of the corporation, with a view to discussing them with Pierre S. du Pont. Regarding the Dayton situation, I said:
I believe that considerable time has been lost in the development of the air-cooled car through lack of appreciation, undoubtedly on the part of all of us, that certain fundamental facts do not, to any extent, subscribe to the contentions of Mr. Kettering—that everybody in General Motors must be sold on the details of his proposed car. I believe that if he had developed a car and demonstrated its performance or it had demonstrated its performance in the hands of independent observers and that he had left the production of the car to others as a principle, that we would have been farther ahead. I think we have made a mistake in putting it all up to Mr. Kettering and not recognizing his particular and peculiar situation. I believe that the Corporation needs and the industry needs advanced engineering. We are not going to get advanced engineering from a mediocre mind such as the average of our engineers compared with that of Mr. Kettering. Advanced engineering always, like advanced everything else, brings down upon it the discredit of ridicule of minds who cannot see so far. For all this reason, such engineering must be demonstrated in such a way that the facts must be accepted rather than theories. I do not think that there would have been any trouble in Oakland or that any changes would have been suggested had Mr. Kettering waited until he had a car that would demonstrate a reasonable satisfactory performance. I fear that the way this is working will result in the loss of a great many of the ideas that we need so badly and can only accept from a man of Mr. Kettering's marvelous ability.
The writing of this memorandum served mainly to mark a turning point in my conduct concerning the air-cooled engine. I began then to pursue a dual policy: first, continuing to support Mr. du Pont's and Mr. Kettering's hopes for the new car, and second, giving support to the divisions in the development of alternate programs of the conventional water-cooled type. And, incidentally, on the side, for a while Mr. Zimmerschied and I looked into a new "Muir" steam-cooling system, which never went into production. While Mr. du Pont had no enthusiasm for alternatives to the air-cooled engine, he did not forbid my taking this position. We simply worked along somewhat different lines. But such a situation between the two leading members of the operating organization was not entirely comfortable and could not last forever.
For the next sixteen months the air-cooled car continued to distract the corporation and to keep its leading officers in a state of tension over the question of what the future product of the corporation was to be.
At the beginning of 1922, the pressure on Chevrolet for the new car was increased, while that on Oakland was somewhat relieved. I took a first step toward the compromise which I felt was necessary to protect the corporation in the event of failure of the new program, and to bridge the gap that had opened between top management and the divisions. As vice president of operations I held a meeting in my room in the Hotel Statler in Detroit on January 26, 1922, with Mr. Mott (car group executive), Mr. Bassett (Buick), and Mr. Zimmerschied (Chevrolet), and reached an understanding that the official air-cooled program for Chevrolet was to be advanced, but with caution. The official program required that the experimental air-cooled "4" models under development at Dayton, "if proper, should be put in production by the Chevrolet Division as of September 1, 1922," that is, just seven months away, though the Chevrolet management had not yet received a test car from Dayton. We agreed, however, "that there is nothing before the Corporation or the Chevrolet Division at the present time to justify the positive conclusion that the air-cooled car should be put into production on the date specified," but that on April 1, 1922, after tests had been made, we could determine a safe program. At the same time we agreed that "a second line of defense should be prepared—this being only a conservative policy." The second line of defense was a parallel effort in the division to improve the existing water-cooled Chevrolet.
As to Oakland, on February 21, 1922, I reported to the Executive Committee and obtained approval to postpone a new schedule for the air-cooled "6," production of which had been canceled. For Oakland then we agreed:
1. To continue its recently established [water-cooled] models for a period of a year and a half, ending June 30, 1923.
2. To eliminate from consideration, as far as Oakland is concerned, the introduction of any air cooled model previous to that date.
3. That any design that Oakland may develop in the meantime will be in accordance with the program of designs already established by the Corporation.
4. That if the economic position of the Oakland models change [s] to the extent that the Division is unable to break even, steps will have to be taken at that time as seem best in view of all the circumstances prevailing at that time.
Since the Research Corporation at Dayton was the only substantial corporation-wide, staff engineering group in General Motors at the time, and it was occupied with the air-cooled-engine experiment, the advanced engineering for the water-cooled models devolved largely upon the divisions. All of the car divisions at that time were in need of advanced engineering staff work to make and keep their conventional cars competitive. At Chevrolet, Oakland, and Olds the need was acute. In other words, the divisions had to attend not only to their principal business of engineering, manufacturing, and selling current models, but to their own staff work on forward engineering as well. Not that matters had been any different in this area previously, but it was our intention to provide for comprehensive engineering staff work for the corporation. From the way in which the Research Corporation chose to operate—that is, as a long-range idea organization centering upon Mr. Kettering's unusual capabilities— it was evident that a gap had opened between his important function and the bread-and-butter type of advanced engineering. I did not then know that a historic distinction was being made in General Motors, but I saw the gap before me and on March 14, 1922, obtained approval of a policy of seeking outside engineering designs for the divisions. This policy would never solve the problem but it would help; many years would pass before the problem was entirely understood and met. Among those whom I consulted at that time was Henry Crane, who later came into the corporation as technical assistant to the president and contributed importantly to the corporation's engineering progress, especially in the design of the Pontiac car. O. E. Hunt had only recently—in October 1921—been brought by Mr. Zimmerschied into Chevrolet as chief engineer and I was not yet familiar with his fine capabilities.
The compromise between the air-cooled and water-cooled developments at Chevrolet was an uneasy one. It was soon accompanied by a change in management. On February 1, 1922, at Mr. Mott's suggestion, William S. Knudsen, who had formerly been Ford's production manager, was brought into the corporation's Advisory Staff and assigned as manufacturing assistant to Mr. Mott. Mr. Knudsen visited Dayton and on March 11 made a report on the air-cooled car, in which he recommended "that the car be put in production at once." He advised me, however, that he meant that production should be started in a small way to test the car both commercially and technically. On March 22 Mr. du Pont obtained the agreement of the Executive Committee to remove Mr. Zimmerschied from his post as general manager of Chevrolet and make him assistant to the president of General Motors; and to appoint Mr. Knudsen as vice president of operations at Chevrolet. Mr. du Pont also proposed to make himself general manager of the Chevrolet Division, while remaining chairman and president of the corporation, and it was agreed.
On April 7, 1922, at the president's request, we officially named the experimental development the "copper-cooled" instead of the "air-cooled" engine. Mr. du Pont wished to differentiate it from other systems of air-cooling. But Mr. Kettering continued to say "air-cooled."
Preparations for tooling up for production of the copper-cooled Chevrolet "4" began, with manufacture expected to st
art about September 15, 1922, at ten cars a day, increasing to fifty a day by the end of the year. The Canadian organization, too, was instructed to develop and introduce a copper-cooled "4." But the spring of 1922 passed without bringing any reality to the new programs. The copper-cooled engine remained under test at Dayton.
Spring sales of cars showed 1922 to be a year of fair recovery, and the Chevrolet "490," underdeveloped as it was in engineering design, was selling again. At a meeting in Detroit in May 1922—at which Mr. du Pont, Mr. Mott, Mr. Knudsen, Colin Campbell, then sales manager of Chevrolet, and I were present—Mr. Mott, with my support, proposed another compromise—namely, to put the new Chevrolet bodies, designed for the copper-cooled car, on the old "490" chassis in the fall, to make certain that we would have something new to sell for the next model year. Mr. Campbell opposed this, saying he was afraid to load up the dealers with "490's" during the winter and then give them copper-cooled cars in the spring of 1923. Again I tried to advance the policy of a dual program, saying: ". . . we should use the copper cooled car as an experiment until April 1st [1923]. Then, if it has become successful and is holding up in the field, increase production on the copper cooled jobs and on August 1st [1923] bring in the car as the sole product of the Chevrolet Division. If the car were not successful, we could continue to manufacture the 490s." The differences thus were laid on the table, but nothing was decided.
The parallel programs and proposals for programs created inevitable tensions in the corporation. Mr. Kettering continued to feel that the divisions were dragging their feet. Oakland, he observed, was now several months behind Chevrolet in the copper-cooled development, and Chevrolet plans, he said, were inadequate. He said in May of 1922 that he was working best with Robert Jack, chief engineer of Olds. Mr. du Pont supported Mr. Kettering's opinion of Chevrolet's plans and in June proposed a stiffening of the copper-cooled program at that division. Since chassis and body changes for the new engine were expected to be complete in the fall, and the change-over then would be concerned with the engine only, he recommended that production of the copper-cooled Chevrolet be scheduled for the forthcoming winter.
In September production had not begun, but the official expectations were optimistic. The plan for Chevrolet was to have a monthly capacity of 30,000 water-cooled and 12,000 copper-cooled cars by March 1923 and to convert the entire water-cooled production into copper-cooled by July or, at the latest, October 1923.
In November, Mr. Kettering noted a lack of interest in the copper-cooled car at Olds as well as at Oakland. I said to Mr. du Pont that I was fearful of the outcome of committing three main divisions to a new, untried engineering car design. Mr. du Pont pointed out to me that the decision had been made by the Executive Committee some months ago "and that the only decision left was the question of a change of front, or the abandonment absolutely of all experiments with water cooled and steam cooled cars." He agreed, however, that there would be no final determination with regard to Chevrolet until May 1, 1923. He then proposed that the Olds program be converted exclusively to copper-cooled.
Mr. du Pont's and my views were then expressed on November 16, 1922, in this compromise resolution of the Executive Committee:
RESOLVED, That the copper cooled program shall be as follows: 1. That the product of the Olds Division as of August 1st, 1923, shall be a six cylinder copper cooled car . . . All experiments and developments of water cooled motors shall be discontinued from this date (November 16th, 1922). 2. The Chevrolet Motor Division will proceed with the development of its copper cooled model cautiously, with a view to determining all factors involved, both commercially and technically, always recognized as being present in the development of any new product, in such a way that the hazard to the Corporation is at all times kept at a minimum. 3. The policy of the Oakland Division will be hereafter determined, but under no circumstances shall the Oakland Company put in production a copper cooled car of any kind or description until the position of the copper cooled car as a type is determined in a broad way by actual experience in the field of a sufficient number of cars, such experience being both of a technical and commercial nature.
Thus at the end of 1922 we were committed exclusively to the copper-cooled program at Olds, to a dual program at Chevrolet, and Oakland was exempted until the new car had proved itself. In December, Mr. Knudsen began to manufacture 250 copper-cooled cars at Chevrolet. The year 1922, like 1921, ended with uncertainty as to what the engineering design of the General Motors product was to be.
At the New York Automobile Show in January 1923 the copper-cooled Chevrolet—chassis and motor—was unveiled. It was priced at about $200 above the standard, water-cooled Chevrolet (now called the "Superior" model), and was the sensation of the show.
The schedule of the Chevrolet Division called for the manufacture of 1000 copper-cooled cars in February, with the monthly rate to increase to 50,000 in October. The only question that seemed to remain at the beginning of the new year regarding the water-cooled car was the exact date on which it should be abandoned. But troubles appeared in the course of production and Chevrolet copper-cooled cars failed to appear in February in large numbers.
Two decisive events occurred simultaneously during the months of March, April, and May 1923. First, we found ourselves in the greatest boom year up to that time in automobile history, the beginning of the industry's first four-million car-and-truck year. Second, difficulties in production had slowed the manufacture of copper-cooled Chevrolets to a walk, and the few copper-cooled Chevrolets that were on the road and being checked by the division produced a large number of reports of troubles, indicating that they were still experimental, unproved, and in need of further development. The question of what to do required no great act of mind. The only Chevrolet we had to sell was the old, conventional water-cooled model. Although it was not a high-performance automobile, even for those days, the Chevrolet water-cooled "Superior" model had been improved and it was a workable car. It ran up record sales that spring.
One could feel that a new era in the demand for automobiles had opened up, and it was imperative that the corporation settle upon its product program for a future that would present itself but once. On May 10, 1923, Mr. du Pont resigned as president of General Motors and on his recommendation to the board I succeeded him in that office. We continued to disagree on the merits of the copper-cooled program, but it was left to me as chief executive officer to make the decisions.
At Olds, in accordance with prevailing policy, all work on the water-cooled car had stopped; the inventory of cars was being sold off at a loss of about fifty dollars a car while the division waited to go into production of the new copper-cooled "6" on August 1, 1923. But the troubles with the copper-cooled Chevrolet clearly threatened the validity of this program.
As president I was of course chairman of the Executive Committee—which was enlarged to include Fred Fisher, head of Fisher Body, and Mr. Mott—and at the first meeting at which I presided on May 18, 1923, I took up the Olds question. I stated the facts about the Olds situation and said: ". . . the continued delay in producing the Chevrolet copper-cooled car is a constant reminder of the uncertainties and the difficulties in engineering and manufacture which would most certainly delay the program and might lead to serious embarrassment to the Olds Motor Works organization at the factory and throughout the world." After a discussion with Mr. Kettering, Mr. Knudsen, and Mr. Hunt, we appointed a committee of three engineers—A. L. Cash, general manager of Northwav, an engine-producing division of General Motors, Mr. Hunt, chief engineer of Chevrolet, and Mr. De Waters, chief engineer of Buick—and instructed them to report on the status of the six-cylinder copper-cooled engine. They presented their report to the Executive Committee at a meeting on May 28, 1923, from which Mr. du Pont, Mr. Haskell, and Mr. Raskob were absent. The report was the main business of the meeting. The engineers said:
That the [copper-cooled "6"] engine pre-ignites badly after driving at moderate speeds in air temperatures from
sixty to seventy degrees. That it shows a serious loss of compression and power when hot, though the power is satisfactory when the engine is warming up from the cold condition.
These major difficulties plus several minor ones which can be reported in detail, if you so desire, lead us to the conclusion that the job is not in shape for immediate production. We recommend that we set it aside for further development and it be left out of consideration as far as immediate production is concerned.
Upon hearing this report the Executive Committee canceled the prevailing copper-cooled program at Olds and instructed the division to proceed with the development of a water-cooled engine that would be able to function on the copper-cooled chassis. We expressed confidence in the principle of copper-cooling as a longer range development, and assigned the copper-cooled "6" engine to Mr. Cash for development at the Northway Division.
At Chevrolet 759 copper-cooled cars had been produced, of which 239 were scrapped by the production men. Of the balance, 500 were delivered to the sales organization. Of these about 150 were used by factory representatives. Something over 300 were sold to dealers, of which about 100 went to retail buyers. In June 1923 the Chevrolet Division decided to recall all of these copper-cooled cars from the field.
On June 26, 1923, in a letter to me, Mr. Kettering proposed to take the copper-cooled engine out of General Motors. He wrote:
We started out to do a perfectly definite thing, which has been done, and it is just the same now as it was a year ago, but in the transition stage certain factors have entered into this, which have confused the thing to the point where, unless things can be clarified, I believe the whole proposition should be dropped. If we cannot get some practical way of commercializing this product, in our own organization, I should like very much to discuss with you the possibility of taking this outside of the corporation and this is a thing which has come up within the last week. I am sure I can get capital and pretty much of an organization to go do this job the way in which I know it can be done.
My Years With General Motors Page 11