are taken with regard to asepsis. A large mass of material is introduced 

 which can have no direct bearing on the experiment and in course of 

 absorption must give rise to a large amount of plastic exudation. Pro- 

 ceeding on the hypothesis that the specific stimulus in this case must 

 resemble the other members of the class of hormones in being a body 

 of comparatively small molecular weight, diffusible and not a colloid, we 

 determined in subsequent experiments to inject such constituents of the 

 tissues experimented on as could be extracted by normal salt solution 

 and filtered through a Chamberland filter. In all our later experiments 

 the tissue, whether placenta, ovaries, or foetus, has been pounded for one 

 hour with sand so as to break up all the cells, and, after the addition of a 

 little salt solution, mixed with Kieselgur ; the resultant powder, which is 

 almost dry, is then subjected in a Buchner's press to a pressure of 300 

 atmospheres. The whole of the fluid of the tissues, containing their 

 soluble constituents, is pressed out. This fluid is sterilised by passage 

 through a Berkefeld filter into a sterilised flask and can then be injected 

 in any quantity either into the peritoneum or subcutaneously without 

 fear of septic trouble. 



The first experiment carried out with this improved technique was 

 on a rabbit eight months old which had lived in a cage in the laboratory 

 since it was 14 days old. This animal received 15 injections spread over 

 a period of 17 days. The injections consisted of the fluid parts of the 

 whole of the viscera of 66 foetuses, varying in age from the fourteenth to 

 the twenty-fifth day. The day after the fifteenth injection the rabbit was 

 killed. It was found that fluid could be expressed from the nipples, which 

 were distinctly enlarged, and on reflecting the skin of the abdomen all the 

 mammary glands were seen to have grown to the size found in a rabbit 

 about eight days pregnant. On microscopic section the gland was seen 

 to consist entirely of proliferating ducts which were lined with an 

 epithelium about three cells deep, presenting numerous mitotic figures. 

 The aspect, in fact, both macroscopic and microscopic, was that of a 

 gland incited to grow by the normal stimulus of pregnancy. 



The question then arose whether one could not, by increasing the 

 number of the injections and the duration of the experiment, carry on the 

 hypertrophy of the gland to the point of formation of secreting alveoli. 

 As the subject of our next experiment we had a rabbit which had been in 

 the laboratory four weeks. The general aspect of the rabbit and the 

 undeveloped condition of its nipples showed it to be a virgin. This 

 rabbit received 24 injections, which were spread over five and a half 

 weeks, an interval of ten days intervening at the end of the first week, 

 during which the animal received no injections in consequence of the 

 temporary failure of our supply of pregnant rabbits. During this time the 

 animal received the fluid extract of the viscera of 160 fetuses. Towards 

 the latter part of the experiment the animal improved considerably in 

 condition, the growth of hair on its abdomen became very active, and 

 the nipples enlarged and yielded a colourless fluid when pressed. On 

 killing the animal the day after the last injection the mammary glands 



