1884.] 107 [Hyatt. 



theory grants that these pits were intermediate forms between 

 true tubes and tracheae. The tracheae of Peripatus might thus be 

 supposed to have a common origin with the incurrent tubes of 

 Porifera and to have retained the original irregularity of their 

 distribution to a certain extent. This would appear to be a nat- 

 ural corollary of our former statements, but as will be seen 

 farther on it is not permissible in view of the probably homoplas- 

 tic origin of the coelomic sacs, and the evidently secondary origin 

 and late appearance of tubes and pores in the Coelenterata. 



The embryos of Echinodermata, as may be seen by the publica- 

 tions of various authors (Mem. Mus. Comp. Zool., vol. ix, no. 3), 

 do not close the blastopore, but this becomes converted into the 

 anus, while the stomodeum is developed from a lateral invagina- 

 tion. This invagination is similar to the osculum of Porifera in 

 the ascula stage in so far as it is not coincident with the blasto- 

 pore and is ectodermic. The same homology seems to obtain in 

 other types which close the blastopore and develope a stomodeum 

 by an independent invagination as in Paludina, and Sagitta. 1 The 

 ingrowth of the rim of the blastopore in Actinozoa to form an 

 actinostome is, therefore, due to a fusion between the primitive 

 stomodeum and the blastopore and not to any natural tendency 

 of the blastopore itself to become invaginated. In other words the 

 primitive stomodeum is not necessarily apart of the blastopore, but 

 an independent invagination which may or may not coincide with 

 it. Lankester's and Sedgwick's objection, that we cannot imagine 

 the exchange in position of the posterior and anterior ends of closely 

 allied forms which is implied in the theory of the independent ori- 

 gin of the stomodeum, does not seem to us well founded. There is 

 no greater difficulty in admitting the exchangeability of character- 

 istics in an antero-posterior, than in a transverse direction. The 

 two ends in most embryos are quite as similar structurally as are 

 the two sides, and an exchange is not at all incredible provided it 

 is supposed as taking place either in embryo or in primitive undif- 

 ferentiated forms. 2 According to Sedgwick's hypothesis (Quart. 

 Journ. Micr. Science, 1884, p. 43) metameric segmentation as found 



1 See list by Balfour, Comp. Emb., vol. ri, p. 281. 



2 If Van Wijhe's discoveries are sustained, this proposition will acquire additional 

 support, and a basis for the theory of antero-posterior symmetry will be established 

 equivalent to that which we now have for bilateral symmetry. 



