448 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



Although the position of the plates as just described appears to be the normal, 

 many variations occur. 



Bury believed that the orals ai"e formed in the vicinity of the left, the basals 

 about the right coelomic sack ; but Seeliger found that the first basal without excep- 

 tion appears so far foi-wai'd that it lies entirely outside the region of the right 

 coelome, and that the fii'st oral appears without any reference to the left coelome, 

 on the contrary reaching with its dorsal and anterior branches far within the 

 region of the right coelomic sack; and there are other displacements which seem 

 not to harmonize with Bury's conclusions. 



But it is significant that at the time of the appearance of the calyx pieces the 

 two coelomic sacks have already abandoned their original position, and that the 

 orals and the basals are formed not in a different antimere of the bilateral embryo, 

 but both in the same antimere. In their horseshoelike arrangement about the chief 

 axis their relations are similar to the i-elations of two metameres. 



Seeliger finds that the calyx plates at first are always separated from the 

 ectoderm by two or three layers of mesenchj'me cells, rarely by only a single 

 layer. Later they move nearer the surface, and their needlelike processes may 

 then be easily demonstrated within the ectoderm. 



Originally the meshwork of the calyx plates, which is finer than that shown by 

 Thomson in Aj^tcdon bifida, lies all in a single plane and does not repeat the 

 curvature of the surface of that portion of the body under which the plates lie. 

 Commonly the basals and orals are about equally large, but the plates of either 

 series may be larger than those of the other. The course and the thickness of 

 the bars of the meshwork may vary in different ways. 



Each plate begins in the same way. Two small calcareous needles appear from 

 which soon branches are given off at an angle of 120° that quickly increase in 

 length, join with similar branches to form again a single bar, and then repeat 

 the process. Later the original bars of the meshwork thicken, the spaces between 

 them become smaller, and the whole aspect of the plate becomes more sievelike. 

 In larva3 raised in an aquarium the increase in the thiclaiess of the plates is slower 

 than in larvae from mothers taken from the sea. 



Seeliger was imable to find a case where two or more originally separated 

 calyx plates later became merged into a single element. 



Anterior to the basals, deeper and nearer the central axis, lie the extremely 

 small beginnings of the infrabasals. In some specimens Seeliger was able to find 

 only three or four, more rarely only two; but he believes that five is the true 

 number, and that in these specimens the whole number had not j^et been formed. 

 However, in some cases he found three infrabasals arranged as described by Bury. 



The usual arrangement of the five infrabasals is in the form of a horseshoe 

 open ventrally stretching over the dorsal half of the embryo. The two lateral 

 infrabasals, right and left, lie quite near each other, and this naturally suggests 

 that the two larger lateral infrabasals described by Bury have arisen through 

 the coalescence of two originally separate points of origin. 



The first trace of a calcareous deposit Seeliger found to be in a cavity in a 

 group of mesenchyme cells, mostly pear-shaped, with readily recognizable globular 



