i897. ^RE THE ARTHROPODA A NATURAL GROUP P m 



Mr. F. W. Hutton's letter opens a question which has engaged 

 my attention for some time, and I think that his fears for the Arthro- 

 poda are fully justified by the facts. I have long felt that, while the 

 term Arthropoda is a very convenient one as expressing a certain 

 kind of structural modification, it does not imply a phylogenetic 

 relationship among the forms included in it. In short, the Arthropoda 

 had not an arthropod common ancestor, and jointed legs cannot be 

 regarded as an absolute criterion of relationship. 



The position of Peripatus is closely associated with this question. 

 Its claims to be included among the Arthropoda are too well known 

 for it to be necessary to recapitulate them here, though we shall see 

 that one of them — the presence of tracheae — is of doubtful value. If we 

 admit Peripatus as a primitive arthropod — an offshoot not far removed 

 from the ancestral stem — the ancestral stem can only be that of the 

 Myriopoda and Insecta. It is quite impossible to insert Peripatus into 

 the ancestry of the Crustacea, and it is equally impossible to find a 

 place for it in any line passing from the Crustacea to the Insecta. 

 The Arachnida must also be excluded from this series, as any 

 attempt to derive the Arachnida from a tracheate ancestry leads only 

 to confusion worse confounded. Even if we leave the probable 

 ancestry of the terrestrial arachnids from aquatic forms, related to 

 Limulus and the eurypterids, out of the question, we find that the 

 lower and more primitive members of the class are in every case pro- 

 vided with a respiratory apparatus of peculiar and characteristic 

 form, which must, in this case, be secondary and developed from 

 tracheae. This being the case not only with the lower orders, such as 

 Scorpionidea and Pedipalpi, but the lower members of the Araneina 

 having also no tracheae, any attempt to construct a phylogenetic tree 

 starting from tracheate forms ends in failure. We must therefore 

 admit that tracheae have been developed twice along two independent 

 lines, and this, it may be noted, weakens the claims of Peripatus 

 somewhat. 



The arachnid question is, however, somewhat of a side issue, as 

 the Crustacea alone are sufficient to show us that, if we admit 

 Peripatus, we must break up the Arthropoda and depose the jointed 

 leg from its present high place. It must be regarded as a product of 

 convergent evolution or homoplasy. Another important point in this 

 connection is noted by Mr. Hutton, namely, that the compound eyes 

 of Insecta and Crustacea are in this case also a product of convergent 

 evolution. The resemblance between them is very close, but I hope 

 soon to be in a position to show that it is only superficial. 



If, on the other hand, we take advantage of the weakening of the 

 claims of Peripatus, due to the necessarily double, and therefore, 

 possibly threefold origin of tracheae, and cast it out from among the 

 Arthropoda, I do not think that we gain much. By doing so we 

 necessarily seriously weaken the absolute position of the jointed leg ; 

 for, though not elaborately articulated, the legs of Peripatus approach 



