190 TORRE Y 



cilia. Tlie prominence tliat is formed at this time persists un- 

 til the time of gastrulation, when the embryo becomes spher- 

 ical again. 



At first the cilia on the prototroch are uniform in size, but, 

 during the differentiation of the trochophore, there appear two 

 bands of longer cilia — one at the upper edge which beat actively, 

 and a narrower one on the lower edge which hang down and 

 move slowly and indefinitely (Text-Fig. 8, A). Between these 

 two rows the shorter cilia ^ are retained. The long flagella, 

 borne entirely by the rosette, are about 20 in number and when 

 the embryo is actively swimming are carried stretched out in 

 front, and bunched closely together, quite as in a pilidium larva. 

 When the animal is at rest the flagella wave about slowly. In 

 the trochophore of about twenty-two hours they are replaced 

 by very short inactive cilia (Text-Fig. 8, A). These long apical 

 flagella occur frequently in annelids and lamellibranchs and per- 

 haps may be regarded as a primitive feature. They have been 

 described in Nucula, Yoldia (Drew, '99) and Dreissensia (Meisen- 

 heimer, '01) and in Eupomatiis, Nereis and Amphitrite. 



b. Variations in Cleavage. — Of great interest in the cytogeny 

 of Thalassema are certain variations in cleavage which occur 

 with surprising frequency. As has already been indicated, the 

 bilaterality of the trochophore is normally first foreshadowed 

 by the formation of the cross, the posterior protoblasts dividing 

 equally and bilaterally, while the anterior divide unequally and 

 spirally — the smaller product lying next to the rosette (PI. I, 

 Fig. 3). The normal bilateral cleavage in the posterior cells is 

 caused by a slight reversal in the direction of the spindle in the 

 d quadrant which lies leiotropically (PI. I, Fig. 2) instead of 

 dexiotropically, as in the other quadrants. However, in a certain 

 number of embryos this bilateral division does not take place 

 and we find a complete reversal to a radial type (Text-Fig. 3, D). 

 Furthermore, the posterior cells divide unequally exactly like 

 the anterior. The intergirdle cells, too, bud off rudimentary 



'Conn ('86) thought that the broad band entirely disappeared and was replaced 

 by a single row of prae-oral and post-oral cilia. A careful examination of living and 

 preserved material, however, has failed to show any evidence that this is the case. 



