This case involves the validity of the Rent Regulation for Housing for the New York City Defense-Rental Area1 which was issued by the Price Administrator on October 8, 1943 and took effect on November 1, 1943. It established March 1, 1943 as the maximum rent date. The New York City Defense-Rental Area consists not only of the five boroughs of the City of New York but also of the counties of Nassau and Suffolk on Long Island. The parties, however, have concentrated their attention upon the regulation as it operates in New York City in view of the fact that the city contains more than 90% of the housing accommodations in the Defense-Rental Area.2 We shall likewise confine our consideration to the regulation as it applies to housing accommodations in New York City.
The complainants are owners of housing accommodations located in New York City which they rent to tenants. They are, therefore, subject to the provisions of the regulation. They filed with the Price Administrator a joint protest against the regulation, which the Administrator denied. They seek to have this court set the regulation aside upon the ground that it is invalid because the maximum rents which it fixes are not generally fair and equitable in that they do not afford the owners of rented housing accommodations in New York City a proper return. They also assert in support of this contention that the Administrator in fixing the maximum rents prescribed by the regulation did not take account of relevant factors of general applicability which required the establishment of maximum rents at a higher level. In order to meet the statutory standard that the maximum rents which the Administrator fixes must be “generally fair and equitable,”3 the complainants urge that those rents must allow the rental industry a fair return on its investment and must accord it profits equal to those which it had earned in normal times prior to the impact of defense activities. They assert that the maximum rents established by the regulation operate so that the industry receives neither a fair return on its investment nor its historical return.
Turning to the complainants' first contention, namely, that the regulation *985•prevents the industry from receiving a fair íeturn on its investment, we find that the record is devoid of any data as to the industry’s investment in the real estate affected by the regulation. The complainants appear to assume that present fair value may for this purpose be treated as the equivalent of original investment. We do not think that such an assumption is justifiable. It is common knowledge that fortunes have been built from comparatively small investments in New York real estate as a result of the unearned increment in values which has accompanied the growth of the metropolis over the years.4 Moreover the only evidence of present fair value which the complainants offer is the assessed value for purposes of taxation. But values established by assessments made for tax purposes may be quite inappropriate for rate making purposes. Bonbright, The Problem of Judicial Valuation, 1927, 27 Col.L.Rev. 492; Missouri Rate Cases (Knott v. Chicago, B. & Q. R. Co.), 1913, 230 U.S. 474, 33 S.Ct. 975, 57 L.Ed. 1571; Wilson v. Brown, Em.App.1943, 137 F.2d 348.
The complainants assert that by New York law assessed valuation and fair value must be identical. Whether the New York statutes and cases cited in support of them so hold is of little significance for our present purpose because the admitted facts are to the contrary. For example there is evidence in the record that real estate in the Borough of Manhattan was sold upon the open market between 1937 and 1943 at prices ranging from 63% to 83% of its assessed valuation for tax purposes. The complainants have thus failed to show either the original investment in or the present fair value of the rental properties subject to the regulation. Since the burden was upon them to do so5 we cannot hold that the regulation is not generally fair and equitable because it prevents the rental industry generally from realizing a fair return upon one or the other of those bases.
The complainants’ second contention is that the regulation is not generally fair and equitable because it prevents landlords from realizing the profits which they normally enjoyed before the impact of defense activities. The Administrator concedes the appropriateness of this test of the validity of the regulation and indeed urges that it is the only practicable way to measure the fairness of the return which landlords will realize on their investment. This court has upon several occasions given tacit approval to such a test. In Madison Park Corporation v. Bowles, Em.App.1943, 140 F.2d 316 we discussed the meaning of the standard “generally fair and equitable” when applied to the maximum rents established by a rent regulation. We concluded from the legislative history of the Act that Congress intended (page 324 of 140 F.2d) “to permit business a yield or profit sufficient for its sustenance in a state of efficiency and for reasonable development and expansion” and that this objective “usually will be attained by permitting profits which the industry was willing to accept prior to defense activities.” In Equitable Trust Co. v. Bowles, Em.App.1944, 143 F.2d 735 we treated as significant upon the contested issue as to whether the maximum rents were generally fair and equitable the fact that the overall profits under rent control were above prewar levels.
The test of the historical return has been applied by the Administrator in price control also. By the use of the so-called industry earnings standard he has permitted price increases in order to compensate for those cost increases which an industry could not absorb without impairment of its normal peacetime earnings. Gillespie-Rogers-Pyatt Co. v. Bowles, Em.App.1944, 144 F.2d 361.
The parties, therefore, are in accord that the regulation is invalid if it fixes the maximum rents at so low a level as to prevent landlords generally from receiving the return enjoyed by them in a representative prewar period. They are not in agreement, however, as to the prewar period which should be selected as a basis for comparison. In passing upon the pro*986test the Administrator employed the average results of the years 1939 and 1940 as the basis for comparison. He urges that those are appropriate years for this purpose. The complainants strongly object to the use of the year 1940 in this connection. They say that in 1940 defense activities had begun to bring about an increase in their operating expenses which resulted in a reduced net return since the rate of occupancy was still declining in that year.
Indeed the complainants suggest that 1939 was a year of abnormal depression in the New York rental market. It may well be that in earlier decades landlords did receive rates of return in excess of that enjoyed in 1939. But as we pointed out in Spaeth v. Brown, Em.App., 137 F.2d 669, 670, “Congress clearly authorized the Administrator to stabilize rents at the level at which they stood in the particular area in question on the most recent date which did not reflect increases resulting from defense activities. It did hot intend that all rent control should be suspended until rentals reached the higher levels of an earlier generation.” 1939 was the last full calendar year which could be described as a peacetime year. It, therefore, appears to be the logical year to use for comparison. By 1940 the war in Europe had begun to show its effects in the economy of the United States, so that it would not be accurate to describe it as a normal peacetime year.
It is true that the Administrator has frequently employed the average experience of the years 1939 and 1940 as a base for comparative purposes in testing the validity of rent ceilings. In the past landlords have not objected to the use of the year 1940 in this connection because it was favorable to them. For example, in defense-rental areas such as Kansas City, Seattle, San Diego, and Detroit this was a procedure to which owners of housing accommodations subject to rent control could have only a theoretical objection since in each instance 1940 was a year of sharply rising net rental income-In New York City, on the contrary, the statistics show that the average net rental income went down nearly 5% in that ■year. The New York landlords, therefore, are the first whose operating experience is such as to cause them to raise the issue in this court. We conclude that the year 1939 was the last normal prewar year in New York City and was, therefore, the appropriate period for use in testing the validity of the regulation here in question.
In support of their protests the complainants presented to the Administrator the results of a survey which they had made of 25,930 dwelling units in the city. This survey provided information relating to income, expenses and net income for the calendar years 1940 to 1943, both inclusive. In April, 1944, the Administrator completed a survey which covered the operating experience of 47,514 dwelling units in New York City 6 for the calendar years 1939 to 1943, both inclusive. In the protest proceedings the complainants abandoned their own survey in favor of the more comprehensive April survey of the Administrator, and urged that it demonstrated the soundness of their contentions. The results of the April survey were discussed at length by the Administrator in his opinion denying the protest. Fie concluded that the evidence failed to show that the maximum rents fixed by the regulation were not generally fair and equitable. His conclusion was grounded not only upon the survey figures but also upon an estimate of the results of operations for one year under rent control which he made upon the basis of the survey figures, adjusted for factors which he believed would be applicable in 1944.
In November, 1944, after the complaint had been filed in this court, the Administrator completed a second survey of the operations of 48,646 dwelling units in New York City.7 The results 'of this survey were admitted as additional evidence in this case. The survey included operating figures for the calendar years 1939 to 1943, both inclusive, for the year ended June *98730, 1944, and for the periods of six months ended June 30, 1943, and June 30, 1944. The November survey thus presented actual operating figures for six months under rent control and for a year which included eight months under rent control. The housing accommodations surveyed comprised more than 4% of the rental apartments of New York City, a sample which the complainants concede was sufficiently large to be statistically adequate. Following the November survey the Administrator revised his prior estimate of operating results for 1944 in the light of the actual figures for the first half of that year as disclosed by the survey.
Much time in argument was spent by the complainants in an attack upon the original and revised estimates and by the Administrator in justification of them. We do not think it necessary to enter this discussion, however, since the November survey has provided us with actual operating figures for a year which includes eight months under rent control as well as for a half year entirely under rent control. We think that these figures showing the results of actual operations under rent control are sufficient to enable us to determine the validity of the regulation and that in view of their presence in the evidence but little attention need be given to estimates of future results which under other circumstances, it might have been necessary for us to rely upon. It is true that we do not have before us actual operating figures for the whole year 1944. But a comparison of the actual results for the six months ended on June 30, 1944 with those for the year ended on that date indicates the presence of a trend favorable to the landlords.
We have no doubt that the Administrator in pursuance of his investigatory function will continue at appropriate intervals to survey the actual operating results in this most important defense-rental area and that he will test the continued adequacy of the maximum rents fixed by the regulation in the light of the figures which he thus obtains. For the purposes of this case, however, we shall confine our attention to the actual results shown by the November survey. We accordingly turn to the figures disclosed by that survey.
For all the accommodations sampled in the entire city the November survey gives the following figures:
*988The survey figures indicate that the net operating income of New York City landlords for the year ended June 30, 1944 was but 95.48% of that which they enjoyed in the year 1939. It shows that for the last six months of that year, a period entirely under rent control, the net return had increased to a rate equal to 99.32% of 1939. The November survey thus demonstrates that at least during a twelve months period which included the first eight months of rent control the industry viewed as a whole was realizing a return substantially less than that which it had enjoyed in 1939, the prewar year which we have indicated to be appropriate as a normal base period for purposes of comparison. It also indicates that during the last six months of that period the net return, although substantially better than for the full year, had not reached the prewar level. Accordingly if the statistics which we have quoted are to be taken at their face value there appears to be merit in the complainants’ contention that the regulation is invalid because it prevents landlords from realizing their prewar net return. The statistics must, however, be examined in the light of the problem which confronted the Administrator in imposing rent control in the area. New York City is a vast metropolitan community containing within its borders approximately 13% of all dwelling units registered in the United States.8 While there are many different types of rental housing accommodations in the city the authorities are agreed that they fall broadly into three categories, which we may describe as substandard, medium priced or competitive and luxury accommodations.
The substandard apartments are variously described as old law, cold water, walkup or tenement. These, as the names imply, are substandard housing accommodations which have survived from an earlier era and which are occupied by the lowest income groups. At the other end of the economic scale are the so-called luxury apartments designed and priced for the highest income groups. Between these two extremes lies the category of medium priced, highly competitive, comparatively modern housing accommodations. While, as is so often true, sharp lines are difficult to draw, the evidence indicates that those housing accommodations which rent for less than $30 per month constitute the first or substandard group, those which rent for $30 per month or more but less than $100 constitute the second or competitive group, and those which rent for $100 per month or more constitute the third or luxury group.9
The Administrator’s November survey classified its results in categories according to the average rentals of the apartments in the individual buildings studied. Thus it furnishes us with the operating results for those rental apartments located in buildings in which the average monthly rental of all apartments in the building was less than $30 and it reports separately similar information for apartments in buildings in which the average monthly rents represent various higher rental ranges. We are, therefore, able from .the results, of the survey to examine the characteristics of rental apartments in the various rental ranges in order to determine whether any of these ranges exhibit peculiar characteristics which set them apart from the others and which call for distinctive treatment in the matter of rent control.
When the figures are segregated for the three major categories of housing accommodations which we have already mentioned the following results are disclosed:
*989
*990Of equal significance to our inquiry are the statistics as to the percentage of occupancy for the entire city and for each of the three groups just described. The figures which have been presented to us represent the ratio of the net actual rental income received to the total rental scheduled, expressed in percentage form. They are as follows:
Percentage of Occupancy, 50,101 Units, New York City
Year ended June 30,
6 mos. ended June 30,
Month of June
1939 1940 1941 1942 1943 1944 1944 1944
All Units 93.14 92.22 91.04 92.15 94.60 97.64 96.44 97.65
Rentals under $30 per month 92.49 92.40 91.03 89.10 86.82 87.13 87.84 88.43
Rentals from $30 to $99.99 per month 93.49 92.36 91.50 92.95 ' 95.73 97.57 98.73 98.72
Rentals of $100 and over per month 91.01 91.03 87.43 88.51 92.60 95.96 97.94 97.59”
In the light of these data let us consider the three categories of accommodations in order to determine whether they-do have distinctive characteristics of which we must take account. Examining the data with respect to the substandard group we are at once struck by the fact that the occupancy of this type of housing accommodation has very substantially decreased since 1939 and that this reduced occupancy has continued under price control in spite of the increasing general demand for housing which has been reflected in a greatly inci eased occupancy rate for the city as a whole. The reason for this phenomenon ia clear. With the constantly increasing earning power which the residents of New York City have enjoyed as a result of defense activities more and more of the tenants of these substandard apartments have left them for more modern and higher priced accommodations.
The statistics undoubtedly show that the owners of apartments in the substandard group are not enjoying the net income which they received in the prewar year 1939. Indeed their operating income has been reduced by more than 25%, a very substantial amount. There is no reason to think, however, that the owners of these substandard apartments have been prevented from realizing their prewar rate of return by the imposition of rent ceilings. It is true that they have experienced an increase in operating expenses. But their reduction in net operating income is due primarily to the fact that their accommodations are not presently in as much demand as they were in the prewar years. For the statistics demonstrate that if the substandard group enjoyed an occupancy rate
equal to the average for the entire city on June 30, 1944 their increased rentals would more than offset their increased expenses so that they would be enjoying net operating income in excess of that realized by them in 1939.
With a vacancy rate substantially in excess of 10% the competition for tenants in the substandard group would make it economically impracticable for landlords in that group to increase their present rents even if the rent ceilings established by the regulation should be raised. It is obvious, therefore, that the increase of 10% in the maximum rents which the complainants seek would not afford the owners of these apartments any substantial relief.
When we turn to the statistics relating to the competitive group of medium priced apartments we notice that the operating expenses of this group also have risen' steadily since 1939. We observe, however, that their occupancy rate has greatly increased during the last two years and that the resulting increase in actual rental income received has fully offset the increase in operating expenses. Thus the figures show that for the year ended June 30, 1944 the owners of apartments in this medium priced competitive group realized 99.11% of the net income which they had enjoyed in 1939 while for the first six months of *9911944, a period wholly under rent control, they received net income at the rate of 102.56% of the 1939 figure. It would thus appear that if the maximum rents were increased 10% as urged by the complainants the owners of rental apartments in this group would be afforded a net return very much in excess of that which they enjoyed in the prewar period.
When we consider the third group comprising the so-called luxury apartments a different picture is presented. The characteristic which distinguishes this group from the others is that during the years since 1939 there has been a continuous decline in the scheduled rentals for these apartments. This is explained by the fact that these high priced apartments arc necessarily occupied only by the wealthiest class of tenants and that many members of this class because of greatly increased taxes during the war period have found that they could no longer afford to occupy apartments in this rental range. Widespread and continuous reductions in rental rates were, therefore, necessary in order to attract and hold tenants.
The luxury group in common with the others has experienced a continual increase in operating expenses. Unlike the substandard group it has also experienced recently a substantial increase in the rate of occupancy. But due to the reduced rentals which were in force on March 1, 1943 and which were accordingly frozen by the regulation, this increase in occupancy has not resulted in increasing the landlord’s net return to its prewar level. Indeed an analysis of the statistics demonstrates that in the case of this group increased occupancy, even though it should reach 100%, could not of itself suffice to produce the net operating income of 1939.
In view of the increased occupancy which the luxury group is presently enjoying it can hardly be suggested that an increase in maximum rents would not be effective to enable the owners of apartments in this group to realize a better return from their properties. On the contrary it is clear that the regulation is actually operating to prevent the owners of this group from realizing from their apartments net income at 1939 levels.
To summarize: Rental housing accommodations in New York City fall into three distinct categories, which may be described as substandard, medium priced or competitive and luxury. The statistical data as to operating experience disclose that the owners of the medium priced competitive housing units are enjoying under rent control a rate of return comparable to that of the typical prewar year 1939 but that the net return of the owners of the substandard and luxury housing units has fallen below that level. It appears that while a horizontal increase in maximum rents would afford relief to the owners of the luxury housing units it would not enable the owners of the substandard units to increase their rate of return. Moreover in the case of the medium priced group an increase in the maximum rents would result in an increase in rental income which would permit the owners of these accommodations to obtain a net return substantially above the prewar level. It is, therefore, clear that a horizontal increase in maximum rents for all rental housing accommodations in New York City, as contended for by the complainants, would be contrary to the purposes of the Act.10 It is equally clear that a denial of any increase would result in one distinct segment of the industry being prevented from realizing its prewar return. We think that the very statement of this dilemma suggests its own solution.
Although it is true, as a general rule, that a regulation must be judged by its effect upon the industry as a whole, there may be occasions when an atypical segment of the industry may call for distinctive treatment. See Adams, Rowe & Norman v. Bowles, Em.App.1944, 144 F.2d 357; Heinz v. Bowles, Em.App.1945, 149 F.2d 277. Upon the same principle the validity of the regulation here in question must be determined separately with respect to its effect upon each of the three distinct groups into which the rental housing accommodations of the New York City Defense-Rental Area are found to fall.
It would be wholly unrealistic to hold that the validity of the regulation must be determined by considering the net result of operations of the rental industry of the New York City Defense-Rental Area as a whole and upon the basis of average figures for the operations of all landlords therein. Since the three distinct segments of the industry, as we have seen, *992have had quite-diverse operating experiences, such an average necessarily presents a wholly theoretical picture which does not reflect conditions as they actually 'exist in the defense-rental area. ' We accordingly reject the city-wide average figures of the November survey as a basis for testing the validity of the regulation. This makes it unnecessary for us to consider the question much discussed in argument as to whether the city-wide average figures shown by the survey were statistically accurate or whether they should have been weighted in order to give effect to the results for each rental range in strict proportion to the frequency of its units in the area.
We turn, then, to the second of the complainants’ two principal contentions. Conceding that March 1, 1943 was an appropriate maximum rent date, they nevertheless contend that the Administrator erred in failing to adjust the maximum rents of that date for factors of general applicability which had arisen since 1939, the last prewar year, and which called for higher maximum rents. The factors, upon which they rely are the increases in operating expenses and the reductions in rental rates which took place during that period. Insofar as either of these factors has resulted in a general reduction of the net operating return below the normal prewar level it must be regarded as a relevant factor of general applicability which under the Act is required to be taken into account in fixing maximum rents.
The results of the November survey indicate, as we have seen, that for the substandard and competitive groups the March 1, 1943 maximum rentals established by the regulation have not operated to deprive landlords of 'the opportunity to receive a net return at the 1939 level. In the case of the competitive group increased occupancy has offset increased expense so that, according to the latest figures in evidence, net returns are actually át or slightly above the 1939 level. In the case of the substandard group it has been the impossibility of securing enough tenants for these undesirable accommodations and not the amount of the maximum rents which has prevented landlords from securing a net return at the 1939 level.
The luxury group also has been subject to the factor of increased operating expense. In addition this group has suffered during the war years a reduction in rental rates which amounts to approximately 10% of the 1939 rental. We have seen that this reduction has taken place during the war as a result of a factor directly related to the war effort. Increased wartime taxes upon the high income group which tenant these apartments had the effect during the earlier war years of substantially reducing the number of available tenants for housing accommodations within the luxury rental range. In order to compete for and retain tenants for these high priced accommodations landlords resorted to the rent reductions to which we have referred.
The statistics show that since the imposition of rent control the rate of occupancy in the luxury group has increased to 97.59%. This largely increased occupancy has not been sufficient to offset the factors of increased expense and reduced rentals to which the group has been subject. It does indicate, however, that the existing maximum rents are operating to prevent the landlords of this group from securing a net return from their properties equivalent to the 1939 rate.
It thus appears that with respect to the luxury class of housing accommodations in New York City there existed prior to the maximum rent date a factor of general applicability to the class, namely, a general reduction in the income of the tenants of the class brought about by high wartime taxes, which compelled landlords as a group to accept lower rents for their accommodations. To the extent that this factor prevented realization by the landlords of the luxury class as a group of a net return at the prewar level we hold that it was a relevant factor of general applicability which should have been taken into account by the Administrator in fixing the maximum rents for this group of housing accommodations. Since, as the evidence shows, the reduced rentals resulting from the influence of this factor are still operating to prevent the landlords of the luxury group from realizing their prewar rate of return, it follows that in the absence of some suitable adjustment the regulation must be held not to be generally fair and equitable in its application to that group.
One further contention of the complainants should be referred to. They urge that *993some allowance should be made for deferred maintenance of rental properties. This contention was made primarily in connection with their attack upon the Administrator’s estimate of the results of operations for the year 1944. The complainants also urged that in determining the adequacy of the return upon their investment allowance should be made for deferred maintenance of their properties. For reasons already stated we do not consider either of these contentions. We do not think that the complainants have proved that the item of deferred maintenance affects in any substantial way the actual operating results which are shown by the Administrator’s surveys and by which we have tested the validity of the regulation. The Administrator points out that in his survey expenditures for replacements and renewals were not included as items of operating expense but were eliminated from the figures for all the years studied. While there doubtless has been some deferment of essential maintenance work during the war period the complainants have not shown the extent of it. It is equally true that landlords have continued to expend large sums for essential maintenance and repairs during this period while at the same time materially reducing their expenditures for painting, decorating, renewing floor coverings and other competitive maintenance and replacement items which can be omitted in a period of maximum occupancy.11
Having considered all the contentions of the parties we conclude that the regulation here under attack is generally fair and equitable and, therefore, valid in its application to the substandard and medium priced classes of housing accommodations in the area. We think, however, that it has been shown not to be generally fair and equitable in its application to the luxury class of housing accommodations which rent for $100 or more per ■month. The evidence before us indicates that in order to be valid in this respect the regulation must be modified so that it will not operate to prevent the owners of housing accommodations in the luxury class generally as a group from securing net returns comparable to the prewar net return of the group as measured by the operating results of the year 1939.
It should be borne in mind that it is the housing accommodations which actually rent for $100 or more per month rather than the buildings in which they may be located for which the modification must be made. It is true that the statistics which were put in evidence were necessarily based upon the operating results of buildings and that some of the buildings with average rentals within a given range did include a few apartments outside the specified range. The statistics nevertheless sufficiently demonstrate that the factors which after 1939 depressed the rents of apartments renting for $100 or more per month did not have the same effect upon the rents of apartments in the medium priced group. Accordingly it would violate the purposes of the Act to increase the maximum rents of apartments renting for less than $100 per month even though they were located in an apartment building the average rents of which were above that figure. Moreover it is quite clear that if as the result of the modification of the regulation the maximum rents of apartments in the luxury group are increased the results will be reflected in the net operating income of the apartment buildings in which they are located even though those buildings also contain apartments in the medium priced group for which no adjustment is made. The form which the modification should take, however, whether by way of a change in the rent ceilings applicable to the class, the inclusion of a suitable adjustment provision, or otherwise, is a matter which the Act commits to the judgment of the Administrator. We accordingly intimate no opinion on that question.
A judgment will be entered setting aside the Rent Regulation for Housing in the New York City Defense-Rental Area insofar as it is applicable to housing accommodations for which under the Regulation maximum rentals of $100 per month or more have been established unless within thirty days from the date of entry of the judgment the Regulation has been modified so that it will not operate to prevent the owners of such housing accommodations generally as a group from securing a net return comparable to the net return received by owners of such housing accommodations generally as a group in the year 1939.
8 F.R. 13,914.
There are 2,412,524 dwelling units in the Defense-Rental Area, of which 2,218,-372 are in New York City.
Emergency Price Control Act of 1942, § 2(b), 50 U.S.C.A.Appendix, § 902(b).
See Myers, History of the Great American Fortunes, 1936 Ed., Part II.
Montgomery Ward & Co. v. Bowles, Em.App.1943, 138 F.2d 669; Madison Park Corporation v. Bowles, Em.App. 1943, 140 F.2d 316, 326.
The April Survey covered the operating experience of 1,583 buildings with a total of 49,067 rental units, of which 1,553 were store units and 47,514 dwelling units.
The November Survey covered the operating experience of 1,642 buildings with a total- of 50,101 rental units of which 1,455 were store units and 48,646 dwelling units.
In 1940 New York City liad 2,218,372 dwelling units. Out of a total of 14,476,-005 rental dwelling units in the United States registered with the Office of Price Administration on June 30, 1944 there were 1,946.219 units registered in New York City alone as of April 30, 1944.
Of New York City’s 1,872,755 tenant occupied and vacant dwelling units for which monthly rents were reported in the 1940 census 581,383 or approximately 31% were in the $30 or under rental range, 1,-226,915 or approximately 66% in the $30 to $99.99 rental range, and 64,457 or approximately 3% in the $100 or over rental range.
Compare Safeway Stores, Inc., v. Bowles, Em.App.1944, 145 F.2d 846.
Compare Chatlos v. Brown, Em.App. 1943, 136 F.2d 490, 496.