This is an appeal from a judgment denying the appellant a refund of corporate income and excess profits taxes levied for the years 1952 through 1956.
In November of 1960, the appellant elected to have the provisions of Section 613(e) (2) and (4) (F) of the Internal Revenue Code of 1954 applied to its taxable years of 1952 through 1956.1 Section 613(c) (4) (F) provides that the cost of mining in the case of calcium carbonates shall include the cost of treatment processes applied prior to the introduction of the kiln feed into the kiln. This section establishes the “prekiln cut off point” and excludes from mining costs any subsequent process, including the kiln burning process. The appellant, in pressing his claim for a refund, contends that certain manufacturing and marketing expenses were wrongfully denied treatment as mining costs.
FACTUAL BACKGROUND
Appellant is a West Virginia corporation engaged in the business of producing and selling finished Portland cement. During the years in question, it owned three quarries of calcium carbonate rock deposits and operated a cement manufacturing plant within 50 miles of each location.
Appellant, in its integrated operation, extracts calcium carbonate rock from the quarries and ships it by truck or rail car to crushing machinery for size reduction. The crushed rock is stockpiled in the vicinity of appellant’s cement plants, as are such purchased raw materials (additives, such as iron ore, pyrite cinder, slag and blast furnace flue dust), which are added to the calcium carbonate rock to provide the requisite iron content of cement. The crushed carbonate rock and the purchased additives are ground together in ball mills and reduced to a fine size. The ground product is then conveyed either to homogenizing silos for dry blending2 or mixed with water in agitating tanks to produce a wet slurry.3 Wet or dry, the raw mix is known as “kiln feed”. The chemical composition of the kiln feed varies according to the particular kind of cement to be made. The kiln feed is introduced into inclined rotary kilns where it is burned at a high temperature and emerges as “cement klinker”. This klinker is ground together with gypsum to an extremely fine-sized powder. That powder is the finished cement, which is stored in silos at the plants pending shipment. The additives were made necessary by reason of the iron deficiency in appellant’s calcium carbonate rock.
Appellant’s total raw grinding and dry and wet mixing costs were incurred for labor, supplies, power, repairs and depreciation of equipment. The costs were incurred both with respect to the processing of appellant’s rock and the processing of the purchased additives. The portion of such costs attributable to the processing of the additives is reflected in the ratio of the weight of the purchased additives to the total weight of the raw mix used in producing cement. On appellant’s financial books and records, the cost of stockpiling and han*506dling the purchased additives was treated as part of the cost of the additives.
In the course of its business, appellant incurred overhead administrative expenses, including such items as salaries of executives and supervisory personnel, legal fees, rental and office expense, general dues and subscriptions, telephone and postage costs, local tax payments and pension and welfare funds. These costs were not directly identifiable with either the mining or the non-mining phase of appellant’s business. Also incurred were plant overhead costs, such as operating offices, store rooms, laboratories, yards, grounds, auxiliary buildings and the services of engineers, technicians, clerical aid and maintenance crews. Appellant allocated 50% of the plant overhead to the kiln burning process and the remaining 50% to the process of grinding cement, klinker and gypsum.
In the marketing of the product, a variety of selling expenses are incurred consisting primarily of the salaries and expenses of salesmen and other sales personnel. Appellant did little direct advertising, but was a member of the Portland Cement Association, to which it paid dues, and which widely advertises the uses and benefits of its cement. To market the product, appellant offered its customers discounts predicated upon prompt payment. These discounts varied depending upon the plant involved. Thus, for example, on sales from its California plant, appellant offered a discount of 20 cents per barrel if the invoice was paid by the 10th of the month following the month of sale. On sales from its Texas and Ohio plants, appellant offered a discount of 10 cents a barrel if the invoice was paid within 15 days after the date of the invoice. This discount practice was apparently uniform throughout the industry and substantially all of appellant’s customers sought and were allowed the discount as offered. The discounts were subtracted from gross sales to arrive at net sales.
Substantially all of the customers orders were filled from the storage silos at the respective plants, although a distribution warehouse was maintained at Paramount, California for the purpose of making rapid deliveries of bulk cement to customers in the greater Los Angeles area. The value of the inventory at the warehouse included the cost of transportation thereto in the taxpayer’s trucks from the manufacturing plant at Victorville. All of the costs of storing and warehousing the cement were treated as manufacturing costs and direct costs of sale. The appellant did not segregate the cost of loading cement in bulk from the cost of loading bag cement and it realized an overall profit on all sales of cement whether in bulk or in bags. It did not separately state the bag premium on its invoices to customers. In the California operation, in addition to the use of hired carriers, the taxpayer employed its own fleet of trucks to deliver cement. All transportation costs were paid by the appellant and treated as part of the total cost of sales. The cost of transportation was not separately stated on the invoices or separately paid by the customers. This was true whether the delivery was made in the appellant’s trucks or in hired carriers.
During the years in question, there were no sales by appellant of kiln feed to independent purchasers, nor were there any markets for kiln feed which would establish representative market or field prices that a taxpayer could have obtained in its area of operation.
The trial court found that the costs of packing and loading cement, of operating the Paramount distribution warehouse and transportation thereto, and of transporting cement to customers, were related solely to specific quantities of finished cement and were directly identifiable with the non-mining phase of the taxpayer’s operations. Moreover, the trial court found that the salaries and expenses of taxpayer’s salesmen were wholly attributable to sales of the taxpayer’s manufactured end product, finished cement, and were directly identifiable with the non-mining phase of the taxpayer’s business. As for the tax*507payer’s advertising costs and trade association dues, the court found that they were related to the taxpayer’s sales of cement and were of benefit to the taxpayer’s non-mining phase. Generally, the trial court found that appellant’s post-manufacturing costs were all productive of income in the course of the taxpayer’s operations and that the taxpayer’s practice of discounting its list prices was a trade discount, a conventional price-making device resulting in a reduction or readjustment of the actual selling price to customers and was not a cash discount to obtain a prompt payment. Finally, the court found that good accounting practice required allocation of plant overhead to all functions carried on at the location of the plant or under plant supervision, including the packing and the loading process.
ISSUES
There are two principal issues involved in this appeal: (1) should certain subdivisions of Treasury Regulation § 1.613-3,4 adopted in July, 1968, be applied retroactively to the tax years involved, 1952 through 1956, and (2) various assignments of error dealing with alleged erroneous treatment of expenses by the trial court in computing the gross income from mining, upon which the depletion deduction is based. The ultimate purpose of each of these assignments is to increase the gross income from mining by varying one or more of the factors involved in the formula.
THE PROPORTIONATE PROFITS COMPUTATION
On the issues before us, it is necessary to determine what portion of the gross income is allocable to mining and non-mining. This allocation must be made in order to determine the allowable deduction for depletion under 26 U.S.C. § 611. The proportionate profits method of computing gross income from mining is reflected in the following equation:
MINING COSTS/TOTAL COSTS x Gross Sales = Gross Income from mining.
(1) Relying upon Helvering v. R. J. Reynolds Tobacco Co., 306 U.S. 110, 59 S.Ct. 423, 83 L.Ed. 536 (1939), the appellant contends that the new reg*508ulation, Sec. 1.613-3 cannot be applied retroactively to the years in question. Notwithstanding appellant’s rather convincing argument that the regulation should not be given retroactive effect, the fact remains that the issue has now been resolved against it in United States v. California Portland Cement Co., 413 F.2d 161 (9th Cir.1969), decided since the filing of appellant’s opening brief. The precise regulations were before the court in California Portland. The court there concluded that the regulations were a valid exercise of the secretary’s power and were retroactively applied to five tax years prior to 1960. Helvering v. R. J. Reynolds Tobacco Co., supra, upon which appellant relies, was considered and distinguished by the California Portland court, on the petition for rehearing. Appellant calls our attention to Bloomington Limestone Corp. v. United States, 315 F.Supp. 1255 (S.D.Ind. 1970). In Bloomington, the district court relied primarily on R. J. Reynolds. As previously mentioned, California Portland distinguished Reynolds and held that it was not applicable to these facts. Appellant does not argue that this case should be distinguished from California Portland on the issue of retroactivity but, in effect, argues that California Portland was wrong. We cannot, of course, even if so inclined, overrule California Portland without calling for in banc consideration. This we decline to do.
(2) (a) Transportation Costs. Appellant would eliminate from “gross sales” and “total costs” the cost of transportation. Its argument, not disputed by appellee, is that costs which are non-productive of profits should not be included in the formula since the basic presumption of the proportionate profits method is that each dollar of cost earns an equal share of the total profit. The new regulation, Sec. 6.613-3(e) (2) (iii) (c) (1968), recognizes the overall problem of non-productive costs where purchased transportation is concerned and provides, in substance, that non-productive transportation costs may be deducted from the formula since they do not produce any share of the total profits. The presumption raised by the regulation is that the transportation costs are productive of profit unless the taxpayer can “demonstrate the non-profit character of the transportation services.”
Here, the trial court made a special finding that the appellant used a delivered pricing system and that “transportation costs” were productive of profit for the appellant. The fact that the findings of fact were prepared by the government and approved without change does not destroy their validity. Since the court’s finding on this issue is not clearly erroneous, we cannot disturb it. The fact that the appellant quoted a mill price would not, in itself, be sufficient to overcome the presumption that the delivered price was productive of profit. Appellant cites no authority in support of his argument that the regulation itself is arbitrary, and we find none precisely in point. Presumptions of this type quite properly place the burden on the taxpayer.
(b) Next, appellant contends that selling expenses should have been allocated between the mining and non-mining phases of its operation. The district court found that the appellant failed to sustain its burden of “ * * * proving to what extent, if any, the expenses of selling its manufactured end product are of benefit to the kiln feed mineral product and the mining phase of the business.” This finding, in our opinion, is in direct conflict with our decision in California Portland, where we said, “* * * [T]he integrated miner, who disposes of a composite of a mined and a manufactured product, and incurs selling expenses on the total product, should allocate these expenses between the mining and nonmining portions of his activities in computing the depletion allowance. Otherwise, the nonintegrated miner would receive a tax advantage not intended by Congress.” 413 F.2d 172. Commenting on the basis of allocation, the court said: “* * * [T]hat the *509selling expenses in the instant case are to be allocated in the proportion that the mining costs bear to the total costs, * * *” id at 172. The conclusion on this point is buttressed by Arvonia-Buckingham Slate Co. v. United States, 426 F.2d 484 (4th Cir.1970).
The appellee practically concedes that the trial court was in error in this finding. It makes a half-hearted attempt to distinguish the facts before us and those before the court in California Portland. The argument is not persuasive. The facts from a legal viewpoint are not distinguishable. Appellee’s argument that California Portland was wrong on this point would call for an in banc hearing. We decline to make such a suggestion.
(c) Next, appellant urges that the purchase price of the pre-kiln additives, plus the cost of grinding and blending the additives with the limestone should have been included as mining costs under the formula. Although appellant stipulated at the time of trial that the cost of the pre-kiln additives did not constitute a “mining cost” within the purview of the formula, it now urges that the stipulation was improvident and that Riddell v. California Portland Cement Co., 330 F.2d 16 (9th Cir.1964) supports a different result. Riddell involved pre-kiln iron bearing additives which were in and of themselves subject to depletion allowances. The pre-kiln additives, in this case, says the appellant, are waste materials from industrial processes which are not subject to depletion allowances and, therefore, the problem of double depletion does not here exist. The short answer to this is that including the purchase price of the additives, as a mining cost, has the overall effect of allowing them to be depleted as calcium carbonate limestone. Certainly, the additives are not limestone.
On the issue of the proper handling of appellant’s costs of grinding and blending the additives with the limestone, the appellant again suggests the importance of Riddell. There, the government conceded that “ * * * [T]he act of physically adding the quartzite and iron ore to the limestone — is a pre-kiln ‘process’ and that the expense thereof is a ‘mining cost’.” 330 F.2d at 18. Even if the government’s concession in Riddell was of importance to us, on the facts here presented we are controlled by California Portland. In the latter, the cost of storing and handling the pre-kiln additives was held not to be a mining cost. The panel in California Portland rejected the contention that storing and handling was analogous to the blending activities considered to be mining costs in Riddell.
We hold that the purchase price of the pre-kiln additives, as well as the cost of grinding and blending this material with the limestone, are not part of the mining costs. The finding in Riddell was based on a government admission. Here, the finding of the district court is not erroneous.
(d) Next, the appellant argues that the trial court erred in finding that the discounts given by it were trade discounts. Appellant says they were cash discounts and, as such, represented a financial expense attributable, in part, to mining. The finding of the district court is entirely consistent with the decision in California Portland. Id., 413 F.2d at 172-173. The finding is supported by substantial evidence which demonstrated that the discounts were given to virtually all customers and were practically identical with the discounts given by appellant’s competitors in the immediate area. The contention is without merit.
(e) Here, appellant contends that there was error in the method by which the court below allocated the plant overhead. It argues that the overhead expense should have been allocated between mining and non-mining costs on the basis of direct labor hours, rather than according to the ratio of mining to non-mining costs. The method employed by the district court in allocating this expense is expressly authorized by the *510Treasury Regulation. See. 1.613-3(d) (1) (iii) (1968). While the method suggested by appellant, if supported by evidence, might have been used by the court if it found it “reasonable under the circumstances”, as provided in the regulations, appellant offered no admissible evidence on the subject. Consequently, the trial court was justified in following the rule outlined in the regulations. The method employed by the district court was not clearly erroneous.
(f) Next, the appellant argues that the cost of bags and bagging should have been allocated between mining and non-mining expense, rather than treating it solely as a non-mining expense. We hold that this issue has been resolved against appellant in California Portland Cement. Id. at 167-170.
(g) Finally, appellant asks us to treat the expenses incurred at the Paramount facility, including the transportation of the cement to the plant, as “selling expenses” and allocate such expenses between mining and non-mining costs; rather than treating such expenses as non-mining costs only. Treasury Regulation 1.613-3(d) (4) (iii) (c) 5 is in direct conflict with this contention. Moreover, appellant is faced with the regulation 1.613-3(d) (4) (i) (1968) providing, among other things, “ * * * [N] on-mining transportation includes the transportation * * * from a mining facility to a non-mining facility, or from one non-mining facility to another. * * * ”
The judgment of the district court is affirmed on all issues with the exception of the issue on the proper allocation of selling expenses. As to the latter, the judgment is set aside and that issue remanded to the district judge for consideration of proper allocation of the selling expenses in the light of the rule stated on the subject in United States v. California Portland Cement Co., 413 F.2d 161, 172, decided subsequent to the decision of the trial court.