delivered the opinion of the court.
This appeal is from the decision of the Patent Office Trademark Trial and Appeal Board (abstract 158 USPQ 681) dismissing the opposition by appellant to the registration of the trademark HYDRO-NOL for “diuretic,” application serial No. 214,501, filed March 19, 1965, claiming first use June 10,1964. We reverse.
The application is to register HYDRONOL on the Principal Register, filed by Atlas Chemical Industries, Inc., a Delaware corporation, and the specimen filed therewith shows that HYDRONOL, at the time of filing was a “new drug” marketed for investigational use only by a division of appellee known as The Stuart Company, Pasadena, California.
Opposer, Geigy Chemical Corporation, now, by merger with Ciba Corporation, Ciba-Geigy Corporation, relies on its prior registration No. 437,263, issued March 9,1948, under the Act of February 20, 1905, for the trademark HYGROTON for “Diuretic remedy for the treatment of oedema of cardiac origin in powder, liquid and tablet form.” The registration recites continuous use since April 25,1947. For legal purposes, the goods are obviously identical. The only question is whether the marks so resemble each other as to be likely to cause confusion, or to cause mistake, or to deceive. (15 USC 1052(d)).
Neither party took testimony. The case was presented below on a stipulation of facts dated April 4, 1967, covering the period from 1960 through 1966 with respect to appellant’s sales and the preceding year as to preparation for sales. The salient facts may be summarized as follows. HYGROTON was adopted by appellant for use on a diuretic having the generic name chlorthalidone in January 1959.1 *974First investigational use thereof commenced March 23, 1959, and an FDA New Drug Application was conditionally approved on March 18, 1960. The first commercial shipment was in April I960. HYGRO-TON has been continuously used since March 23, 1959. Total sales through 1966 in dollars have been $20,772,384 and in tablets 407,547,-100. Advertising inserts in the same period were 1240 and expenditures for advertising of all hinds including samples were $3,274,132.
Priority of use and registration by opposer is not disputed.
Appellant asserts and appellee does not deny that Reg. No. 437,263 of March 8, 1948, was brought within the Trademark Act of 1946 under section 12(c) thereof (15 USC -1062(c)) by publication in the Official Gazette of December 6,1960, affidavits under sections 8 and 15 were filed and accepted January 19, 1966, and the registration was renewed in 1968. Originally granted to a predecessor, the ownership of the registration now resides in opposer-appellant.
The basis of the board’s decision dismissing the opposition is to be found in its statement that the marks of the parties are
* * * readily distinguishable when spoken and, in our opinion, they do not closely resemble one another in appearance, irrespective of whether they are presented in handwritten form or in block type.
The board appears to have been concentrating on a side-by-side comparison. When so compared, we might come to a similar conclusion but what the statute prescribes as the test is whether the marks so resemble one another as to be likely to cause confusion or mistake, and this requires us to consider, among other things, the fallibility of memory over a period of time, not merely whether one can distinguish the marks at a given moment.
Given even a moderate interval of time, and even though these are prescription drugs, we do not feel that physicians who prescribe these two diuretics would be immune from confusion either as to the products or their source. The prefixes HYDRO- and HYGRO-sound very much alike and convey almost the same meaning. There is also an obvious similarity in the last four letters of the marks, -ONOL and -OTON, namely, tjhe same number and the same placement of the two “Q’s” and the same number of letters, not only in those suffixes but also in the marks as a whole. This may be dissection, but its purpose is only to point out the similarities in the marks as a whole.
We think that a doctor who, perhaps, had recently been introduced to HYGROTON or HYDRQNQL as a diuretic, not prescribing diuretics to Ms diverse patients every day, on encountering the other product a week or two later or searching his memory or the index of *975tlie “Physicians’ Desk Reference” (PDR) and finding both trademarks there might not be sure which was which, particularly since both are diuretics and therefore have the same physiological function, generally speaking. The question is not whether a physician is careful, as he normally is, but how long he remembers clearly when confronted with confusing similarity. We feel there is a reasonable likelihood that, by reason of the similarity of the marks, he would be confused.* Since reliance may be placed on fallible memory of the sound or appearance of a trademark, as distinguished from a generic name, we feel there should be a clearer distinction than exists here.
Both sides have, of course, cited many prior decisions on other marks which, as usual, we find of little help. In the end, it is a subjective opinion. Our considered judgment that there is reasonable likelihood of confusion and mistake as to both product and source could be supported by citation of many precedents in which there were even greater differences between the marks than exist here. It goes without saying that if there were any [4] doubt about likelihood of confusion, it must be resolved in favor of the prior user. Appellant has persuaded us that the board erred and we therefore reverse.
The record shows nothing about the use of the mark from 1947 through 1959 by appellant or its predecessors, but the 1959 use is sufficient to make appellant the prior user.
The 1969 “PDR,” in a list of 63 “Diuretics,” lists four proprietary products whose brand names begin with. “HYDRO-” and one beginning with “HYGRO-,” viz, HYGROTON.