340 PALAEOSTRACA. 



early stages in the development of the germ-band in the Scorpiones, as described 

 by Metschnikoff and others (Vol. iii.). In the latter also are found at first an 

 anterior and a posterior region of the body, a few small free segments, arising 

 successively, being cut off from the posterior region and intercalated between 

 the two regions. Similar stages are found in the Araneae. We must, however, 

 bear in mind that the free segments of this germ-band are later drawn into 

 the head, and are evidently represented in the Trilobites by the segments of 

 the glabella. These cannot, therefore, be homologised with the free thoracic 

 segments of the Trilobites. A certain similarity is, however, brought about by 

 the presence of a large posterior section from which the trunk-segments arise 

 later. 



Barrande pointed out that most species of Trilobites in which metamorphosis 

 occurs belong to the older strata of the Bohemian Silurian, while, in the more 

 recent strata, young stages of Trilobites are much less frequently found, although, 

 in these same strata, the number of species is richer, and the conditions for 

 the preservation of these delicate forms are also to some extent favourable. 

 Barrande therefore seems justified in conjecturing that metamorphosis was 

 perhaps replaced by direct development in the later and perhaps more specialised 

 Trilobites. 



II. Xiphosura. 

 Limulus polypkemus is the only Xiphosurid whose ontogeny has 

 as yet been investigated.* The metamorphosis of L. mollucamis is 

 still unknown, hut it should he mentioned that Willemoes-Suhm 

 (No. 31) believed that a pelagic Crustacean larva obtained off the 

 Philippines by the Challenger, which resembled the Cirripede larvae 

 (p. 209), could be referred to L. mollucamis. He, however, changed 

 his opinion later, and classed the larva among the Cirripedes. 



1. Oviposition, Cleavage, and Formation of Germ-layers. 



The eggs of L. polypJiemm are laid on the seashore in holes dug 

 in the sand half-way between high and low water mark (Kikgsley, 

 No. 14), or even near the former. L. rotandicauda and L. mollu- 

 canus, on the contrary, do not deposit their eggs, but carry them 

 about attached to the swimmerets. The eggs, when laid, are enclosed 

 in a very thick leathery membrane composed of several layers, which 

 has been described as the chorion, and is perhaps not formed in the 

 ovary itself (Dohrn, No. 11), but in a special portion of the genital 

 ducts. As the embryo increases in size, this membrane bursts, bo 

 that, in the later stages, the embryo is covered solely by the later 

 formed blastodermic cuticle. 



For information as to the first ontogenetic processes in the egg of 

 Limulus wc arc entirely dependent on the short accounts of OsBORN 



• L. longiqwnus has also recently been investigated by KlSHlNOUYE [Zool. 

 . 1 1, No. 869), but we are not here able to give his statements in detail 

 [see App. Lit. on Xiphosura, No. II. See note, p. 343.] 



