COKPUSCULAR ELEMENTS OF THE LYMPH. 525 



brane (goldbeater's skin) and introduced tinder the skin of a 

 living rabbit. At the end of twelve hours, a few small leuco- 

 cytes and granulations had made their appearance ; at the 

 end of twenty-four hours, the fluid had become somewhat 

 opaque and contained a large number of leucocytes and 

 granulations ; and at the end of thirty-six hours, the fluid 

 was white, milky, and composed almost entirely of leucocytes 

 and granulations. The leucocytes, which were examined also 

 by Robin, presented all the characters by which these cor- 

 puscles are ordinarily recognized. These experiments were 

 repeated with more than forty different specimens of fluid 

 from blisters. 



The experiments were then varied in order to show the 

 influence of the membrane and of the composition of the 

 blastema upon the development of leucocytes. By modi- 

 fying the membrane in which the blastema was enclosed, 

 it was found that the corpuscles were rapidly developed 

 in proportion to the activity of the osmotic action. "When 

 thick animal membranes were used, their development was 

 slow, and in some instances did not take place at all. There 

 was no development of leucocytes in a clear blastema enclosed 

 in a sac of caoutchouc or in glass tubes hermetically sealed ; 

 and from this it was concluded that the osmotic action was a 

 necessary condition, and that the mere heat of the body was 

 not sufficient to develop these corpuscles, even in an appro- 

 priate blastema. The influence of this constant molecular 

 movement is in striking contrast to the conditions of absolute 

 repose which are so essential to the formation of crystals 

 from ordinary saline solutions. 



One of the most interesting points in these experiments 

 is connected with the influence of the composition of the 

 blastema upon the development of leucocytes. It was found 

 that these bodies were never developed in a blastema in 

 which the fibrin had been coagulated. Experimenting with 

 two liquids, the only difference in their constitution being 

 that in one the fibrin had been coagulated by repeatedly 



