SPICULES OF SYNAPTA AND AUKIGULARIA LARVA. 489 



add considerably to the picturesque effect of tlie whole. As 

 the spicule increases in size, the nuclei of the portion of 

 syncytium concerned in its formation distribute themselves 

 more uniformly over its area. Apparently very few, if any, 

 nuclei are subsequently added to those initially present in the 

 internal cluster which gives rise to the plate-spicule. It must 

 also be remarked that the special internally-situated portion of 

 the syncytium which produces the plate becomes almost quite 

 separated from the rest of the syncytium enveloping the 

 anchor as the growth of the plate proceeds, since the anchor 

 and the plate gradually diverge to include the angle pre- 

 viously mentioned (text-fig. 1) ; the syncytia of the anchor and 

 of the plate in fact alone remain continuous at the joint formed 

 between the handle of the anchor and the base of the plate, 

 and this continuity of the two syncytia in this region probably 

 serves to some extent both to keep the two structures in 

 apposition and to render the joint a true joint, i. e. a mutual 

 centre of rotation (see fig. 44, in Study IV, 7). 



To return to the later development of the anchor. Up to 

 the time when the arms of the anchor-bow are but half-grown, 

 the syncytium in this region is, as before mentioned, stretched 

 to form two " patagia," so to speak, on the two sides of the 

 shaft (figs. 16, 18). But on further elongation of the arms 

 the central portions of the patagia apparently give way, with 

 the result that the scleroplasm reniaius in the neighbourhood 

 of the shaft as a thin layer, and away from the shaft as " two 

 elongated strands of protoplasm containing many nuclei 

 which run on either side from the arms of the bow to the 

 handle " (Study IV). When the arms are fully formed, these 

 two strands (figs. 20-24) are thickened at the bow end, 

 forming '^ blobs," and the nuclei at this end of the anchor are 

 almost entirely collected in these two thickenings. There is 

 also in each strand another such collection of nuclei situated 

 in the vicinity of the handle. Nuclei do occur in other parts 

 of the strand, but they are principally aggregated in these 

 two regions ; they also occur, of course, on the shaft, bow, 

 and handle, though not in clusters. Thus the formation of 



