230 Edmund B. JFilson. 



lar or even amoeboid, and I have sometimes observed even trocho- 

 blasts of very irregular form. But this is not the case with most 

 of the isolated cells in Patella and Dentalhim, and I am inclined 

 to accept the probability that the cells in question may really be 

 mesenchyme- and perhaps actually muscle-cells, that have dif- 

 ferentiated in more or less complete isolation from their fel- 

 lows. If this be considered an improbable conclusion, it should 

 be recalled that a trochoblast is probably, to say the least, as 

 highly differentiated as a mesenchyme cell; yet it has been strictly 

 proved that such a cell may undergo its normal differentiation 

 and continue for a time to perform its normally coordinated ac- 

 tivities when completely isolated from the time of its formation. 

 Further research specifically directed to this point will, I believe, 

 give a positive result on this very interesting question. 



5. Summary en isolated cells from the lower hemisphere. 



The evidence derived from these cells is less detailed and com- 

 plete than that derived from the first quartet; but as far as it 

 goes gives the same general conclusion. The isolated >^-macro- 

 mere, yV-macromere or second quartet-cell segments as if form- 

 ing part of a whole embryo, and shows more clearly than do the 

 first quartet cells that not only the form, but also the rhythm of 

 cleavage is maintained (precisely as I showed in the nemertlne) ; 

 for In the cleavage of both the ^s- and the yV-macromere the 

 fourth quartet cell is the last to form. Only the embryos con- 

 taining derivatives of the second quartet produce secondary 

 trochoblasts, namely, those arising from the ^-macromere or the 

 second quartet-cell. While all the embryos close, only those gas- 

 trulate that contain the basal region (/'. e., the entoblast region). 

 All of the three types examined develop one or two feebly cil- 

 iated cells that probably represent cells of the pre-anal ventral 

 ciliated tract. Finally, there is some evidence, though only of an 

 inferential character, that Isolated mesoblast-cells may develop 

 into mesenchyme-cells, possibly into muscle-cells. We may, there- 

 fore conclude that, speaking broadly, the development of cells 

 of the lower hemisphere, like that of the upper, conforms to the 

 mosaic principle. 



