598 HAROLD HEATH, 



quartette, are in contact wiih the tip cells of the second quartette. 

 A considerable distance therefore exists between the groups of primary 

 trochoblasts which is filled to a large extent by the basais, the 

 remainder being completed by the accessory trochoblasts. 



In AmpJiitrite the cell corresponding to the accessory trochoblast 

 is larger than the basal cell, and this latter, owing to its small size 

 and the relatively large development of the primary trochoblasts, 

 becomes crowded back into the angle between the arms of the rosette 

 series. The basais therefore are not in contact with the tip cells 

 (slight contact in one quadrant of Fig. C, c), and the space between 

 the groups of primary trochoblasts is filled with the accessory trocho- 

 blast, and thus the girdle is complete. But it must be noted that 

 the girdle is also completed by the tip cells of the second quartette 

 and it is this girdle that becomes functional, the accessory trocho- 

 blasts being ultimately pushed above the functional prototroch. Thus 

 it happens that in Annelids the cells corresponding to the accessory 

 trochoblasts never enter the functional prototroch. What their fate 

 may be we do not know. In the posterior quadrant of AmpJiitrite 

 a product of the accessory cell is very minute, with a darkly staining 

 nucleus, and owing to these characteristics it was used as a "land- 

 mark". It was never seen to divide, and to me it has much the 

 appearance of a degenerate cell. In one or two of the remaining 

 quadrants a small cell is figured forming a cell corresponding to the 

 one just mentioned. These also, judging from the figures, have a 

 dense and darkly staining nucleus, but whether they degenerate has 

 not been determined. It would be very interesting to discover if these 

 blastomeres do degenerate and are cast out. If they do it would lend 

 much to the view that these cells were once functional locomotor 

 cells, but losing that function they have degenerated as a result. 



As regards the cross. In AmpJiitrite, given smaller trochoblasts, 

 a smaller rosette series and a proportionately larger intermediate cell, 

 we would have as a result of the division of the latter cells the same 

 conditions that exist in IscJinocJiiton] in other words there would be 

 an Annelid cross similar to the one in IscJinocliiton. And conversely: 

 in IscJmocJiiton if the size of the trochoblasts and rosette series were 

 to be increased and the basal cell were proportionately decreased, the 

 conditions as they exist in AmpJiitrite would be realized. The ar- 

 rangement of the homologous cells is practically the same, and the 

 difference in the size of these cells in the Annelids and Molluscs will 

 explain the presence or absence of the cross. 



