Following our reversal (266 F.2d 58, 267 F.2d 802) of a judgment of the district court in favor of Cory Corporation, a Delaware corporation, in consolidated suits for refund of excise taxes collected on sales of air conditioners in 1954 and 1955, the United States Supreme Court, 363 U.S. 709, 80 S.Ct. 1331, 4 L.Ed.2d 1508, reversed this court, and held, at page 711, 80 S.Ct. 1333 that the horsepower test is a permissible one and that the revenue rulings which were in force from 1948 to 1959 were not void. The Supreme Court mentioned, at page 712, 80 S.Ct. at page 1333 that the district court had found that “Among engineers, the horsepower of a motor does not mean its nominal horsepower rating but means the actual horsepower which the motor will deliver continuously under its full normal load”. It pointed out that we did not reach that question nor review that finding in view of our conclusion that the horsepower test was not valid. The case was remanded to this court for consideration of that and any other questions which may remain.
1. Accordingly, on this record, we accept the horsepower test as valid. The question then arises whether the Commissioner, in applying that test, adopted the engineers’ definition of the horsepower of a motor as the actual horsepower which the motor will deliver continuously under its full normal load, or its nominal horsepower rating. We shall refer to these measurements as the “actual horsepower” and “rated horsepower” tests respectively.
Under date of September 26, 1947, the Commissioner, having solicited the views of the air conditioning industry, received a letter from the Air Conditioning and *769Refrigerating Machinery Association, reading, in part, as follows:
“The term ‘self-contained’ is one which has had a broad and somewhat loose connotation as a trade term in the industry. It has been applied to the small, portable air conditioner, and it has been used to describe a thousand-horsepower centrifugal refrigeration unit such as is used for air conditioning purposes. It seems probable that, in writing Section 3405(c) of the Internal Revenue Code [26 U.S.C.A. § 3405(c)], Congress used the term ‘self-contained air conditioner’ in attempting to describe the portable, plug-in electric appliance known as a room air conditioner, similar in general concept to the commonly-used household refrigerator.
“For purposes of taxation, it is desirable that there be available a definition, as exact as possible, of the article subject to tax. The only truly ‘self-contained air conditioner’ being marketed today, so far as we know, is the room air conditioner (room cooler). * * *”
On December 12, 1947, Deputy Commissioner Bliss wrote the York Corporation, a manufacturer of air conditioners:
“ * * * The Bureau has recently had occasion to re-examine its position with respect to the scope of the tax on ‘self-contained air-conditioning units’ imposed under section 3405(c) of the Internal Revenue Code with a view to determining whether it would be possible to formulate a legally sound definition of the term which could be simply applied and which would have uniform application throughout the air-conditioning industry.
“From a consideration of section 3405(c), it appears that the law could, with justification and without regard to other factors, be interpreted to include within its scope any air-conditioning assembly which is so constructed that all of its meehanical components are housed in a single cabinet. To give effect to such an interpretation of the law, the term ‘self-contained air-conditioning unit’ could be defined as follows:
“ ‘A self-contained air-conditioning unit within the meaning of section 3405(c) of the Internal Revenue Code is an encased assembly of a condensing unit with other components which serves as a means for ventilation and for cooling, dehumidifying, cleaning, and circulating air.’
“The foregoing definition does not take into consideration such additional factors as to whether the unit may be installed in the room to be conditioned and may be used without duets for the distribution or return of the conditioned air. One of the principal difficulties in the way of formulating a definition which would give effect to these additional factors and which, at the same time, would be susceptible of fair and easily understood application in the industry is the variety of conditioning units being manufactured and the varied claims made by manufacturers with respect to the functioning and use of the units produced by them. In general, however, it is believed from the information available that these additional factors may be fairly reduced to a test expressed in terms of the rated horsepower capacity of the compressor motor used in the air-conditioning unit. If these additional factors are to be taken into consideration in fixing the scope of the law, the term ‘self-contained air-conditioning unit’ could be defined as follows:
“ ‘A’ self-contained air-conditioning unit within the meaning of section 3405(c) of the Internal Revenue-Code is an encased assembly of a condensing unit with other components which (1) has a compressor motor, the rated capacity of which is 5 h. p. or less, and (2) serves as a means for ventilation and for cool*770ing, dehumidifying, cleaning, and circulating air.’
“The Bureau desires to extend to the industry an opportunity to be heard and to express its views in the matter. Accordingly, the Bureau would appreciate receiving a statement of your views at the earliest practicable date. Similar letters are being addressed to each manufacturer of air-conditioning units of which there is a record in this office.” (Italics supplied.)
Replying to that letter, the York Corporation on January 15, 1948, wrote to the Commissioner a detailed letter, offering a definition of a self-contained air conditioner, within the meaning of § 3405(c):
“ * * * ‘A self-contained air conditioner, within the meaning of section 3405(c) of the Internal Revenue Code, is a factory made encased assembly primarily designed for free delivery of air, and for installation in a window or in front of a window; containing means for moving outside air through its condenser; and having means for ventilation and for cooling, dehumidifying, cleaning, and circulating the air of a room; and having a total motor horsepower of less than one horsepower or a total cooling capacity of less than 10,000 Btu’s per hour at standard American Society of Refrigerating Engineers test conditions as set forth in ASRE Circular No. 16 dated June 1940.’
“We believe the suggested definition is sufficiently broad in its scope to include without exception all self-contained air conditioning units which are now being manufactured. The definition gives recognition to the fact that some units function by means other than electric motor driven compressors. The horsepower limitation was inserted for purpose of facilitating administration, inasmuch as it is not feasible from an engineering standpoint to employ a motor of greater horsepower capacity in a self-contained air conditioning unit.
“ * * * we attended a special meeting called of industry members at which time the problem was discussed fully.
“The definition that we have suggested was unanimously adopted by all who attended, and was formulated with the realization of the administrative problems involved as well as the fundamental question of what constitutes a self-contained air conditioning unit as the term is commonly understood among those who produce such equipment.”
On April 21, 1948, Acting Commissioner Sherwood wrote to Air Conditioning and Refrigerating Machinery Association, referring “to the various conferences held with you and other representatives of the air-conditioning industry,” and stated that “the following definition has been formulated to indicate the position of the Bureau as to the type of units properly to be included within the scope of” § 3405(c):
“A self-contained air-conditioning unit, * * * (3) has a total motor horsepower of less than one horsepower or a total cooling capacity of less than 10,000 BTU’s per hour at standard American Society of Refrigerating Engineers test conditions as set forth in ASRE Circular No. 16 dated June, 1940.”
We are convinced that the foregoing facts established that the Commissioner had been given the assurance of the industry that a rated motor horsepower test of less-than-one-horsepower, to use the words of York, “is sufficiently broad in its scope to include without exception all self-contained air conditioning units which are not being manufactured”, since “it is not feasible from an engineering standpoint to employ a motor of greater horsepower capacity in a self-contained air conditioning unit.”
It can hardly be presumed that the industry intended to mislead the Commissioner, knowing as it must have, that ac*771tual horsepower of a motor might vary from less to more than one horsepower depending on the particular motor and the test employed. The industry must, therefore, have made its recommendations on the rated horsepower which motor manufacturers were assigning to their motors at that time. It is clear, and the court below found, that it is the practice of the industry for air conditioner manufacturers to list the horsepower of a household-type unit as the rated horsepower assigned by the motor manufacturer1 and not the actual horsepower.
That the Commissioner was thinking in terms of rated horsepower is borne out by the fact that in his letter of December 12, 1947 to York Corporation he suggested that units with a “rated capacity” of five horsepower or less be subject to the tax. While the rulings as finally promulgated omitted the reference to rated capacity, failure to set forth any other test or procedure for determining horsepower, although a test was carefully designated for determining the cooling capacity of absorption types, suggests the Commissioner intended to follow the recognized practice of the industry and apply the rated capacity test. The rated horsepower test gives the necessary uniformity which the industry and the Commissioner were seeking in order to facilitate and simplify administration of § 3405(c).
It is obvious that the Commissioner accepted the nominal ratings used in the industry. He did not even solicit reports on tests which would reveal actual horsepower under existing engineering standards for each unit on the market. Neither was such information tendered to him by the manufacturers in the air-conditioning industry. This corroborates our conclusion that taxability was determined by horsepower ratings adopted by the manufacturers which appeared on the plates affixed to the units sold and was not determined by actual horsepower output.
Moreover, prior to issuing the 1954 ruling which revised the 1948 ruling in respects not material here, the Commissioner held a conference with representatives of the Air Conditioning and Refrigeration Institute. We have examined the stipulated facts as to said conference and subsequent correspondence and do not find that they in any way show a change in the interpretation of the 1948 ruling as applied to the horsepower of self-contained air conditioners.
Our interpretation of the 1948 and 1954 rulings makes for a practical enforcement of § 3405(c), which ultimately imposes a tax upon the retail purchaser of an air conditioner. Whether or not such an individual is to bear a tax upon his purchase can readily be established by an inspection of the manufacturer’s horsepower rating tag affixed thereto. If the converse were true and only a test for actual horsepower of the motor in operation would reveal whether the conditioner is taxable, a purchaser would be practically unable to make such a determination.
In view of the foregoing, we conclude that in his 1948 and 1954 rulings the Commissioner adopted the rated horsepower test and made a rating of less-than-one-horsepower the line of demarcation between taxability and nontaxability. The reference of the Commissioner to the rated horsepower capacity of the compressor motors used in the air conditioning units indicate that his rulings were meant to conform with the prevalent views of the industry. While it is apparent that there was not complete unanimity in the industry as to these views,2 there is sufficient basis for *772interpreting the rulings of the Commissioner as a formal adoption thereof.
Because of what we have said, we hold that finding 53 of the district court, in several inseparable parts, is clearly erroneous. It reads:
“The evidence fails to show that it was the intention of the Commissioner of Internal Revenue or of his subordinates, in drafting the 1948 ruling and the 1954 ruling, to refer to nominal horsepower ratings as prescribed by NEMA standards or as selected by individual motor manufacturers, nor does it appear that the Commissioner or his subordinates intended to make taxability turn upon such nominal ratings rather than upon actual horsepower output. * * * ”
2. We now deal with matters which may properly be included in “any other questions which may remain”, 284 F.2d 768.
The amount directly involved in the case at bar is $29.93, being the total of two excise tax payments collected on the sales of two air conditioning units sold in 1954 and 1955, as referred to in findings 3 and 4 of the district court. The taxes were based on § 3405(c) of the Internal Revenue Code of 1939, 26 U.S.C.A. § 3405(c), which imposed a 10% tax on “[s]elf-contained air-conditioning units.” To qualify for bringing their suit for refund of taxes, plaintiffs alleged in their complaints, and it was stipulated in the district court, that Keith McKy and Thomas Carlin would have testified, if called as witnesses, that they each bought such a unit and executed a written consent to a claim for refund, copy of which was attached to the complaint. It appears from the separate opinion of Mr. Justice Clark herein, in which the Chief Justice and Mr. Justice Black joined, that these purchasers were a lawyer and an accountant of plaintiffs.
The district court found:
“5. Defendant’s successor, the present District Director of Internal Revenue, Chicago, has assessed and proposes to assess against Cory, Mitchell and Pearce, substantially more than one million dollars in federal excise taxes upon the sales of air conditioning units similar to those described in paragraphs 3 and 4.
“6. The air conditioning units described in paragraphs 3 and 4, being those described in the complaints, and the similar units described in paragraph 5 are sometimes hereinafter collectively referred to as plaintiff’s air conditioners.
“7. It has been heretofore stipulated by the parties and found by this court that the present consolidated civil action is a prototype or test case to determine whether plaintiff’s air conditioners are subject to excise tax.”
That court also concluded as a matter of law, inter alia:
“3. By reason of stipulations and orders heretofore entered herein, this civil action brings before this court for determination the issue of whether air conditioning units similar to those described in the complaint, on which federal excise tax assessments have been made and are proposed to be made by defendant and his successor as District Director of Internal Revenue, Chicago, are subject to federal excise tax.”
In the majority opinion, the Supreme Court said, 363 U.S. at page 711, 80 S.Ct. at page 1332:
“There is much said in the briefs and in oral argument about this case as a test case. It is said that taxes *773on the sale of about 50,000 units turn on this decision. * * * ”
On remandment, we are therefore immediately confronted with the question of whether plaintiffs have shown that the two cases now before the court with their small money content, are actually “prototype or test” cases so as to justify the extension of the effect of the result herein to other cases not before the court. The record does not show that taxes on the sale of the “about 50,000 units” have been paid. If they have not been paid, there is no basis for a suit for refund such as was brought in the cases at bar. Even if the taxes were paid in reliance upon the validity of the revenue rulings, the litigation now before the court fails utterly to show a similar right of recovery by taxpayers in the cases not before the court, because there is no showing that any consents have been obtained by said taxpayers as required by 26 U.S.C.A. § 6416(a) (3), 1954 Code.
Neither by allegations of facts in the complaints or in the stipulations, nor by evidence introduced, did plaintiffs establish in the district court that, as a matter of fact or of law, the cases at bar are prototypes of the cases involved in the sale of about 50,000 units which are not before the court. Similarity has not been shown. It therefore becomes our duty to limit the effect of the judgment of the district court to the parties named in the judgment entered on June 11, 1958, in which Cory Corporation was plaintiff and Ernest J. Sauber, as former district director of internal revenue, Chicago, was defendant.
We find that, for the reasons hereinbe-fore set forth, findings 5, 6 and 7 of the district court, insofar as they find that the present action is a prototype or test case to determine whether plaintiffs’ air conditioners are subject to excise tax, are clearly erroneous and we set aside said findings in that respect. We hold that the said civil action brought by plaintiffs is not a prototype or test case and is not to be given effect as such by the district court
The stipulation of the parties to the effect that this is a prototype or test case is not sufficient to support the disrict court’s conclusion. While the parties to a civil action, in which only they have an interest, generally may set up for court determination a prototype or test case, in such form as they see fit, public interest in the protection of the federal revenue is an overriding consideration which bars a blanket adjudication of a large total of tax refund claims, based upon test cases privately contrived especially for that purpose. Moreover, this “test” case, which is intended to serve as a release of 50,000 disbursements from the federal treasury, lacks the bona fide consents of buyers which the refund act, supra, requires.
In Swift & Co. v. Hocking Valley Ry. Co., 243 U.S. 281, 289, 37 S.Ct. 287, 289, 61 L.Ed. 722, the court said:
“If the stipulation is to be treated as an agreement concerning the legal effect of admitted facts, it is obviously inoperative; since the court cannot be controlled by agreement of counsel on a subsidiary question of law. * * * ”
Whether or not a large multitude of cases is to be controlled by the case at bar as a prototype or test case is hence a question of law as to which the stipulation of the parties is not binding upon the court. In re Gubelman, 2 Cir., 10 F.2d 926, 929. To the same effect is Davis v. Commissioner of Internal Revenue, 7 Cir., 241 F.2d 701.
This case involves the public revenue to an important extent. In such a case, the parties are not free to stipulate away the rights of the public. This principle is forcibly applied in National Bank v. Murphy, 384 Ill. 61, at page 64, 50 N.E.2d 748, 749, where the court said:
“ * * * No valid stipulation or agreement, as to questions of law or the legal effect of facts, can be made by the parties. * * * Neither can they stipulate the legal effect of admitted facts. * * * ” (Citing *774Swift & Co. v. Hocking Valley Ry. Co., supra.)
And, at page 65, 50 N.E.2d at page 750:
“The decision of questions of law must rest upon the judgment of the court, uninfluenced by stipulations of the parties or their counsel. This is also true as to legal conclusions arising from the facts involved. * # -»»
For these reasons, we determine that the judgment entered by the district court on June 11, 1958 for $29.93 and costs in favor of plaintiff, Cory Corporation, and against defendant, and ordering execution thereon, is to have no effect whatsoever upon other claims or cases, as an adjudication of a prototype or test case.
For the reasons hereinbefore stated, the aforesaid judgment of the district court is reversed.
Judgment reversed.
. Such rating is affixed to plates fastened to the exterior of the compressor housings.
. In the field of manufacturers of electric motors and generators, there is a private trade association known as National Electrical Manufacturers Association, some of whom make motors for plaintiff’s air conditioners. This association’s publications state that the existence of an NEMA standard “does not in any respect preclude any member or non-member from manufacturing or selling products not conforming to the standard”.
The district court found that within the NEMA range for 1 horsepower rat*772ing, as interpreted by motor manufacturers who are members .of NEMA, the proper horsepower rating of plaintiff’s motors under NEMA standards would be 1 horsepower.
The district court also found that motor manufacturers are not required to comply with NEMA standards and are free to depart from them.