No. 553] DISTRIBUTION OF THE CH2ET0GNATHA 37 



necessary to point out that this is an exceedingly impor- 

 tant line of investigation, for, if change of environment 

 and of environed organism are not inseparably linked 

 together, the hypothesis of " natural selection," with its 

 attendant hypotheses of "survival of the fittest," 

 "struggle for existence," etc., are at stake. Ask yourself 

 if it is not a priori impossible for any of these hypothet- 

 ical factors to operate in the formation of species except 

 on the basis of variations in structure which are more or 

 less adapted to the conditions of existence in which an 

 organism finds itself? Again, does not logic demand that, 

 if isolation be a necessary cause of species formation, two 

 similar species must occupy similar but not identical or 

 vastly different environmental complexes, because both 

 could not be equally adapted to the same conditions by 

 virtue of their organic difference nor to radically differ- 

 ent conditions by virtue of their organic similarity? 



Such questions sufficiently indicate the importance of 

 our inquiry regarding the relation between the morpho- 

 logical and distributional characteristics of species and 

 in this connection the key reveals the fact that those 

 species having the most coincident vertical distribution 

 are those having the greatest morphological difference. 

 In other words, when the Chaetognatha of this region are 

 classified in the usual taxonomic fashion, five groups can 

 be distinguished, of which each group contains species 

 having the same fundamental morphological character- 

 istics; but, when classified according to similarities and 

 differences in vertical distribution, the species consti- 

 tuting any one of the five groups are those differing from 

 each other in fundamental distributional characteristics. 

 We have, then, two methods of classification, one of which 

 results in groups of similar morphological but dissimilar 

 distributional species, while the other results in groups of 

 similar distributional but dissimilar morphological 

 species. To illustrate concretely, the groups resulting 

 from each method of classification are tabulated below : 



