Experimental Studies on Germinal Localization. 221 



closely similar external appearance, each is from its first forma- 

 tion definitely specified, irrespective of its connection with its fel- 

 lows. 



4. Development of isolated apical cells, and secondary trocho- 

 b lasts. 



In spite of many attempts, I did not succeed in rearing singly 

 one of the rosette cells; but isolated products of these cells, as 

 well as isolated secondary trochoblasts, were obtained in another 

 way. This was by allowing isolated micromeres of the first quar- 

 tet to continue their development in the calcium-free water, some 

 of the individuals being left undisturbed, others shaken to pieces 

 from time to time by a stream from a fine pipette. In those left 

 undisturbed for 24 hours all degrees of disintegration were ob- 

 served, loose masses being found from which the trochoblasts 

 often had broken away and were swimming about singly. In 

 many such loose masses the characteristic apical cells could often 

 be observed, loosely attached to their fellows at the end opposite 

 the trochoblasts (as in Fig. 30). Those that had been shaken 

 to pieces showed a collection of more or less completely separated 

 rounded cells, among which spherical trochoblasts of two sizes 

 (the larger evidently being the primary, the smaller (Figs. 46- 

 48), the secondary ones) were actively swimming singly or in 

 groups. Among these cells occur forms that are evidently sin- 

 gle apical cells, since they agree exactly in size and structure with 

 those attached to the loose masses referred to above. These cells 

 (Figs. 31-33) are ovoidal in form, and bear on one side the 

 characteristic non-vibratile radiating sensory processes or hairs. 

 There is no possibility of mistaking these cells for either kind of 

 trochoblast, since they are considerably smaller, and the appear- 

 ance and arrangement of the sensory hairs (apart from their im- 

 mobility) is entirely different from that of the cilia. There can, 

 therefore, be no doubt that both the secondary trochoblasts and 

 typical sensory cells of the apical plate may undergo their charac- 

 teristic differentiation when entirely isolated fro?fi their fellows. 

 Mingled with the foregoing cells are rounded, non-ciliated cells 



