306 ME. H, M. BEENAED ON THE 



Hexapodan affinity, chiefly because tliey appear to have a head separated from a thorax 

 of three segments ; this, taken together with the presence of Malpighian vessels, 

 unknown in the Crustacea but characteristic of the Hexapods, was brought forward as 

 evidence. The Galeodidce, therefore, might be expected to be valuable witnesses one 

 way or the other. 



The present attempt to gather the evidence to be gained from the Galeodidse began in 

 the following way: — In 1892* I endeavoured to show that the morphology of the 

 Crustacean A^nis (and of the Merostomata) could be explained by deducing them from a 

 Chsetopod Annelid, which, in adaptation to a new manner of feeding, bent the first 

 segment, with mouth and prostomium, ventrally, so as to push the food iuto the mouth 

 by means of the parapodia. The necessary conclusion, that Liniulns was a primitive 

 Crustacean, brought me involuntarily into the question of the origin of the Arachnida, 

 inasmuch as the supporters of the Arachnidan affinities of Limulus separated the 

 Merostomata from the Crustacea. I therefore set myself the following problem : — Is 

 it possible, by comparing the known Arachnids one with another, to find out, as I claim 

 to have done in the case of the Crustacea, what modifications of the primitive Annelidan 

 segmentation, in adaptation, probably, to some special manner of feeding, gave I'ise to 

 the Arachnidan phylum ? Por I assumed, at the outset, that modifications for better 

 acquiring the materials for growth were the most far-reaching and fundamental in 

 initiating new departures in animal morphology. The Arachnids might still be 

 derivatives of the Merostomata, in spite of the fact that the latter were, as I believe, 

 primitive Crustacea. The only way to settle the question was to ascertain, if possible, 

 how the arrangement of their anterior segments had been derived from a jirimitive 

 ■undifferentiated condition. Here, again, Galeodes promised to yield valuable evidence, 

 on account of its claim to have a head and three thoracic segments. 



The results here recorded were not obtained at once. I had already drafted the 

 description of the anatomy of the Galeodidse iu extenso, and had worked over the draft 

 at least twice without discovering any satisfactory explanation of the morphology of the 

 anterior segments of the Arachnida. The difficulties seemed insuperable, in spite of 

 the apparent simplicity of the results. One clue, however, led to another, each one in 

 turn necessitating considerable recasting of the MS. It is, indeed, with reluctance, 

 under the pressure of other engagements, that I now let it go forth ; for, though I believe 

 that I have solved the problem I set myself, the paper covers so much ground and 

 deals with so many points that I am painfully conscious not only of blanks which I 

 wished to have filled, but of very uncertain handling of many subjects, where I have 

 neither done justice to my own accumulated material nor to the existing literature. 

 I need hardly apologize for errors which, in a work of this extent, are almost 

 unavoidable, except where I have unconsciously overlooked or misunderstood the views 

 and statements of others. 



I have at the outset to fulfil a pleasant duty in thanking the authorities at the Royal 

 College of Science for having placed a table in the Huxley Research Laboratory at my 



* "Tho Apodidae." ' Nature' Series : London, 1892. 



