32 



were found to form a thin sheet extending over almost the whole of the 

 abdomen. Each gland was about five centimetres in diameter and con- 

 sisted of branched ducts radiating from the nipple. The peripheral 

 zone of the gland was swollen and vascular and on section was found to 

 contain alveoli. Both in the alveoli and in the epithelium of the ducts, 

 which was many-layered, many mitoses were seen, showing that cell 

 proliferation was actively proceeding. 



During the last 17 days of this experiment the rest of the bodies of 

 the fcetuses after the extraction of the viscera were finely minced and 

 ground up and boiled with normal saline solution. While boiling the 

 mixture was rendered slightly acid to precipitate the proteids and then 

 filtered. The filtered extract was injected intraperitoneally into a rabbit 

 which had been in the laboratory about two months but from the size of 

 the nipples had evidently been previously pregnant. After the injection 

 had been repeated for 10 or 12 days it was found that milk could be ob- 

 tained by squeezing the nipples. On killing the animal the abdominal 

 wall was covered with a well-marked layer of mammary tissue which 

 contained many alveoli. In some of the ducts there was some indication 

 of proliferation. Both alveoli and ducts contained fluid with fat granules 

 and colustrum corpuscles. Although our failure to obtain a virgin rabbit 

 for this last experiment rendered it impossible to be absolutely certain 

 that the gland tissue had undergone hypertrophy as the result of our in- 

 jections, the occurrence of secretion and the appearances observed in 

 parts of the gland rendered it probable that we were really dealing in 

 this case with a specified stimulus to the mammary gland and suggested 

 that the specific hormone of the gland would therefore withstand boiling. 



In our next series of experiments we employed four rabbits. Two 

 were injected with the extract of the viscera, boiled and unboiled, and 

 two with extracts of the rest of the foetus, also boiled and unboiled. For 

 injection with the unboiled extract of viscera we took a rabbit which we 

 knew to have been previously pregnant. The entire uterus had, how- 

 ever, been removed so as to avoid the possibility of any cooperation on 

 the part of the uterus in the effects of our injections. For the injections 

 of boiled viscera a rabbit was taken which, so far as we could judge, was 

 nulliparous. Of the rest of the foetus the boiled extract was injected into 

 a rabbit which had certainly borne young some time previously. But for 

 the unboiled extract we obtained a rabbit which was nulliparous. Since 

 in this series the foetuses were divided into four parts the injections were 

 smaller than in the previous series. Each rabbit received 22 injections, 

 spread over 31 days, obtained from go foetuses. The foetuses varied in 

 age from the twentieth to the twenty-sixth day. The results of this last 

 series were shortly as follows. In both the virgin rabbits there was 

 hypertrophy of the mammary glands, due chiefly to duct proliferation, 

 though not so marked as in the previous series. Towards the end of the 

 injections a watery fluid could be obtained by squeezing the nipples of 

 either of the two rabbits. In the two multiparous rabbits it was impossi- 

 ble to be certain that any actual growth of the gland had taken place. 



