COMPAEATIVE MOEPHOLOGT OF THE GALEODID^. 38T 



the cocoon. It has been suggested to me by my friend Mr. Pocock tliat tliis was the primitive 

 function of the spinning-mamillie, and that from it all the web-spinning of the Spiders has been 

 secondarily developed. 



XIII. Summary : an Aitempt to Elucidate the Phylogeny of the Arachnida. 



In a comparative study like the foregoing, it has been imjjossible to avoid drawing 

 conclusions as to the relative order in time of the appeai'ance of the different specializa- 

 tions. Some must obviously have preceded others and led up to them. Hence, at the 

 close, we find it possible to arrive at some conclusion as to the primitive ancestral form, 

 and we are in a position to construct a hypothetical form which possesses (1) only those 

 specializations which ai-e common to all the members of the class, (2) these same 

 specializations only in that stage of development which comparison leads us to believe to 

 have been the most primitive. The different specializations of the ancestral form must 

 all be at stages from which the more complicated specializations of the different members 

 of the group can be deduced. 



It is obvious that we can have no clear view of the position of the Arachnida among the 

 Arthropoda until we have arrived at some fairly accurate conclusion as to the common 

 ancestor of the former. It remains to be seen whether the type we can reconstruct on 

 the lines just laid down has any points of affinity with any existing specialized type of 

 Arthropod, or whether we have to seek its nearest relatives in the common ancestors 

 of all the Arthropods, the segmented Chsetopod Annelids. 



The way for such speculations is, however, not altogether clear, inasmuch as a vigorous 

 and long-sustained attempt has been made to demonstrate a relationshij) between the 

 Arachnids and the Merostomata. For clearness of argument I propose therefore, in the 

 following hypothetical reconstruction of the ancestral form, after each section, to show 

 how it bears for or against such relationship. 



Figures 14! (PI. XXIX.) and 12 (Pi. XXXIII.) show the most important characters, 

 external and internal, of our reconstructed ancestral form. This appears to possess most 

 of the specializations common to the group in their least specialized condition. 



1. Number of Segments. — We may faiiiy assume that the common ancestor possessed 

 18 segments {Scorj^io, Thelyphonus), the greater number of segments being evidently the 

 least specialized number. In the majority of the Arachnids, this number is, as a rule, 

 more or less reduced in adaptation to certain specializations which will be mentioned 

 presently ; the full number is only retained where the terminal segments have been 

 secondarily developed into a kind of tail. 



This number agrees with that of the segments of the Eurypterids, from which it is thought Scorpio 

 might be descended. 



2. Fusion of the Segments. — A comparison of the different Arachnids shows that the 

 first fusion embraced the three anterior segments. These were bound togetlier by the 

 distortion of the first segment, which, by the formation of the cephalic lobes, was tilted 

 upward and backward so as to cover the 2ud segment and part of the 3rd (PI. XXVII. 

 fio-. 8). Gideodes and (?) Schizonotus are the only known Arachnids which have remained 



51* 



