CHAETOGNATHA COLLECTED BY STEAMER ALBATROSS. 237 



the other faunas, and so of the whole economic and fisheries situa- 

 tions of the coastal waters on opposite sides of the Pacific at corre- 

 sponding latitudes. Extensive exploration of the Pacific, particu- 

 larly of the coastal waters of Central and South America, is needed, 

 however, to discover the full significance of what is here so clearly 

 indicated, and it is regretted that no chaetognatha are described 

 from these regions. But, in spite of this, it seems probable, from the 

 meager data at hand, that conclusions reached through explora- 

 tions in the western Pacific are largely inapplicable to the waters of 

 the eastern Pacific, and the opposite. Some space is therefore taken 

 at the close of the paper in briefly comparing the Philippine and San 

 Diego chaetognatha. It is hoped this will emphasize the need of 

 more extensive explorations, and that it may add its mite toward a 

 better understanding of the fishery problems of the Pacific Ocean. 



KEYS FOR THE IDENTIFICATION OF THE CHAETOGNATHA. 



Ritter-Zahony (1911) has been the last investigator to thoroughly 

 revise the chaetognatha. He recognizes six genera Sagitta, Ptero- 

 sagitta, Spadella, Eukrohnia, HeterokroTinia, and KroTinitta. Sub- 

 sequently Germain and Joubin (1912) added two more Pseudo- 

 sagitta and EroTinitella. All are probably valid with the possible 

 exception of Spadella and Pseudosagitta, the status of which is baf- 

 fling. Most of the differences given by Ritter-Zahony (1911) between 

 Pterosagitta draco and Spadella cepJialoptera are certainly no greater 

 than that between those species of Sagitta in which the skeletal part 

 of the vestibular ridge is present and those in which it is absent, and 

 this difference is clearly of subgeneric rather than generic value. 

 On the other hand, Conant's (1895) description of Spadella schizop- 

 tera, although fragmentary and wholly unsatisfactory, reveals a close 

 affinity between that species and S. cepJialoptera and at the same 

 time makes the genus to which it belongs unmistakably distinct 

 from Pterosagitta. Furthermore I have seen specimens of neither 

 S. cepJialoptera nor S. schizoptera, and it seems best, therefore, to 

 tentatively recognize Spadella as valid in spite of the fact that the 

 characteristics by which its one well known species, S. cepJialoptera, 

 differs from P. draco seem of subgeneric value. 



The validity of Pseudosagitta is ably discussed by Baldasseroni 

 (1915, p. 101), who holds its single new species P. grimaldi to be 

 synonymous with Sagitta lyra. The differential characters described 

 by Germain and Joubin (1912, p. 6) are certainly such as to suggest 

 this synonymy and I find myself in agreement with Baldasseroni. 



In the following keys seven genera are therefore recognized, of 

 which Sagitta is represented by 23 species, Eukrohnia by three, 

 Spadella by two, and each of the others by one: Pterosagitta draco 

 (see p. 264), HeteroJcroJinia mirabilis (Ritter-Zahony, 1911, p. 42), 



