98 NATURAL SCIENCE. February. 



I offer no opinion myself on the subject, but should like to know 

 how others view the situation. 



Christchurch, N.Z. F. W. Hutton. 



No zoologist can be astonished that the genealogical trees of the 

 animal kingdom, which have sprung up so plentifully during the last 

 few decades, should require serious pruning. I take it therefore that 

 Mr. Hutton's object is to ascertain how those who have made a 

 special study of Arthropoda propose to arrange their genealogies at the 

 present moment. Experience shows that the less one knows about a 

 group, the easier is it to dash off its family tree, while, conversely, 

 the more one knows, the more reluctant are the branches to group 

 themselves in any satisfactory manner. Thus the genealogy of the 

 Arthropoda will not be finally established until each of its component 

 groups has received far more attention than has yet been devoted to 

 it — that is, with this special object in view. 



In giving my own suggested ancestry of the Arthropoda, I should 

 state at the outset, that it is based upon special studies of only two of 

 the groups concerned, the Crustacea and the Arachnida, The 

 detailed arguments on which the following summaries of my 

 conclusions are based, will be found in my " Comparative Morphology 

 of the Galeodidae " [Trans. Linn. Soc, London, 1896.) and in the other 

 papers there cited, and further, in a summary now in the press for 

 the March number of Science Progress. 



(i). There is no difference of opinion in referring all the 

 Arthropoda bacl^ to Chaetopod Annelids, but inasmuch as there is no 

 reason to believe that the Arthropoda are the only derivatives from 

 Chaetopoda, I should not count these latter also as Arthropoda. They 

 are only the soil out of which the Arthropoda developed along a 

 definite line, viz : the specialisation and structural modification of 

 the parapodia for the purposes of prehension and locomotion. 



(2). The lines along which these specialisations and modifications 

 of the parapodia have developed for these purposes are, so far as my 

 researches go, not the same in any two groups of Arthropoda, hence 

 each group has to be deduced separately. The Arthropoda thus 

 cannot be arranged in any single genealogical tree. 



(3). The Crustacea. Regarding Aptis and the Trilobita as the 

 earliest and most annelidan Crustacea, we find that their peculiar 

 specialisation was primarily the bending round of the prostomium 

 ventrally to form a great labrum under which a certain number of 

 parapodia could function as jaws within and around the mouth- 

 aperture, one or more of the anterior pairs being left free to function 

 as sensory organs. The parapodia posterior to the mouth-organs 

 seem ventrally to have repeated the chewing functions of the jaws, 

 while dorsally they developed into complicated swimming plates, 

 retaining the original parapodial gill. The development of walking 

 legs was secondary as is shown clearly by the Trilobita, which 



