818 



SCIENCE 



[N. S. Vol. XXXIII. No. 



sembled the foster mother that he was led to 

 ascribe the result to original ovarian tissue 

 of the foster mother. This assumption was 

 based upon another assumption, namely, that 

 chicks from the engrafted ovaries would pre- 

 serve the characters of the fowl from which 

 the ovaries were obtained. The fallacy of 

 this assumption has been pointed out above. 



Davenport did not use standard varieties of 

 fowls, so far as I am able to determine from 

 his statements. This is unfortunate, as it is 

 obviously impossible to discuss his findings 

 from the standpoint of relationship of donor 

 to host. For example, I have shown that en- 

 grafted ovaries in fowls do not succeed if the 

 stock is too distantly related. 



Davenport states that my results justify the 

 opposite conclusions to those which I have 

 drawn; but since he does not give any rea- 

 sons nor present any evidence for such a con- 

 clusion, it carries no weight other than as a 

 personal opinion. 



Castle and Phillips ask that my esperi- 

 ments be repeated before they accept my inter- 

 pretation of the results. In reply, I ask why 

 they did not employ fowls (chickens) in order 

 to confirm or discredit my experiments. I 

 may say that my first series of fowls, operated 

 on in the summer of 1904, were all lost 

 through lack of proper facilities. The next 

 series, operated on in 1906, were given my 

 undivided attention and furnished the ma- 

 terial for my papers. A larger series operated 

 on the following year with the view of extend- 

 ing the observations and investigating new 

 fields opened up by the successful series, were 

 not productive of results in the direction of 

 permitting the study of offspring from en- 

 grafted ovaries, but furnished considerable 

 information along other lines which is in 

 part presented in my later papers. Success- 

 ful breeding of fowls, as every one knows, 

 demands the fulfillment of certain require- 

 ments in the way of quarters, and facilities 

 for hatching and raising the chicks, and in- 

 telligent attention. As to the first two of 

 these requirements, the third series of experi- 

 ments clearly proves that the quarters and 

 facilities at my disposal, though after a man- 



ner adequate for eight fowls, the number com- 

 posing the second series of experiments, were 

 not adequate for five times this number, the 

 approximate number that were included in 

 the third series. Also, it was not possible for 

 me to give as much time to the third series 

 as to the second. Immediately following this, 

 I made application to the officers of one of 

 the endowed research funds for support in 

 prosecuting the investigation on a much 

 larger scale, which included the employment 

 of a number of species of animals. But for 

 perfectly good reasons the request was denied. 

 Since that time new experiments have been 

 continuously in progress, but they have been 

 designed with a view of keeping within the 

 limits of my facilities. 



I do not propose to enter into a discussion 

 of Castle and Phillips's results in this place, 

 save to challenge their assertion that theirs 

 is the first critical case of successful ovarian 

 transplantation from the standpoint discussed 

 above, on record. This statement I make in 

 view of the fundamental considerations also 

 above stated, as well as from an examination 

 of their protocols. For example, they used 

 mongrel stock. Therefore, any evidence fur- 

 nished by the character of the offspring would 

 be of doubtful value. This is true particu- 

 larly as regards soma influence; and as cross- 

 breeding was not employed, any evidence of 

 soma influence in the offspring would have 

 been obscured by the character of the male 

 parent. 



Also it is not proven that the offspring may 

 not have come from ovarian tissue of the host 

 left in site after operation. Indeed, an inter- 

 pretation of their results from the numerical 

 standpoint, a criterion employed by them in 

 interpreting their results from the Mendelian 

 standpoint, it would be as fair to conclude 

 that in all of their pigs that became pregnant 

 no post-mortem findings are given. And 

 after operation that this was due to incom- 

 plete removal of ovarian tissue. For they 

 state that of the five animals in this group, 

 the results in three were due to ovarian tis- 

 sues generated from the host. Of the two 

 animals left in the successful group, for one 



