EXPERIMENTS WITH FROG'S EGG>. 89 



of graded strengths. The amount of salt is given in percentages 

 although this is not quite accurate, since, in the first place, one 

 per cent., for example, is one gram added to IOO cc. (and not to 

 99 cc.) of water ; and in the second place a certain further de- 

 crease in strength is caused by the water in the jelly around theeggs. 



It is not an easy matter to determine definitely where normal 

 development ceases, for, not only is the development often stopped 

 at different stages in the same solution, but there is also a con- 

 siderable overlapping in the different solutions. I have arbitra- 

 rily selected as the upper limit, the strength of solution in which, 

 although the cleavage may continue for a time, gastrulation does 

 not occur. 



For LiCl alone it was found that gastrulation may take place 

 and a normal embryo appear in a 0.5 per cent, solution. In 

 a 0.55 per cent, solution the blastopore is large, its closure is 

 delayed, and even prevented. In a 0.6 per cent, solution the 

 blastopore appears, but generally no embryo develops ; while the 

 upper limit seems to be about 0.65 per cent, where practically all 

 of the eggs fail to pass beyond the late segmentation stages. For 

 NaCl the upper limit is above 2.0 ; but owing to an accident the 

 upper limit was not accurately determined. 



In a solution of LiCl 0.4 per cent, -f Nad 0.3 per cent, only 

 the late segmentation stages appeared. Thus both solutions, 

 being lower than the maximum for each, produce an injurious 

 effect. In another double solution of. LiCl 0.5 per cent. + NaCl 

 0.5 per cent, a few nearly normal embryos developed, but in a 

 solution of LiCl 0.5 per cent, -f NaCl i.o per cent, only the late 

 segmentation stages developed, or rather a cap of black cells ap- 

 peared at the top of the egg (surrounded in some cases by a 

 band of gray cells, as in lithium larvae that I formerly described). 

 In a solution of LiCl 0.5 per cent, -f NaCl 1.5 per cent, only 

 late segmentation stages developed. This then may be taken as 

 the limit. 



A solution of cane sugar must be quite strong in order to 

 prevent the development of the embryo. Even a 6.0 per cent, 

 solution gives normal embryos. An 8 per cent, delays the 

 closure of the blastopore, and a 10 per cent., while not prevent- 

 ing the appearance of the blastopore, does prevent its closure, 



