Carpenter — Melatiotis/tips between Classes of Arthropoda. ;3-ll 



segments in Limulus, anything like an ancestral standing for that 

 animal is unthinkable ; while the Eurypterids, although primitive in 

 their abdominal segmentation, had the cephalothorax completely fused, 

 and the sixth pair of limbs specialized as paddles. Moreover, they 

 were contemporary with the earliest Scorpions, Howes, in a recent 

 criticism ('02) on the upholders of the Arachnidan affinities of Limulus, 

 has pointed out that the Limuloid type is simj)lified by the Eurypterids, 

 as the modern Scorpionoid type is by Palseophonus. This is just 

 criticism : but it should not prevent us, led by the close correspondence 

 between the segmentation in the primitive Scorpion and in the 

 Eurypterids, from carrying the simplification still farther back, and 

 deri-sdng the Gigantostraca on the one hand, and all the modern air- 

 breathing Arachnids on the other from aquatic Arachnid ancestors, 

 with three free thoracic segments. In the detailed account of the 

 Silurian Scorpion (Palaeophonus) lately given by Pocock ('01), it is 

 especially noteworthy that the post-abdomen is relatively thick, and 

 that the appearance of the ventral sui'face suggests aquatic respii-ation 

 by lamellate gills. 



Bernard ('96) has contrasted the ventral position of the mouth in 

 the Gigantostraca with its terminal position in the undoubted 

 Arachnida, and has founded on this contrast a plea against the union 

 of the former with the latter group. But this argument surely lays 

 too great stress on an adaptive character. It might forbid us to 

 derive Scorpions and other predaceous Ai'achnids directly fi'om 

 Eurypterids, but not to trace both back to a common ancestiy. And 

 fui'ther, if the forward position of the mouth, as in Galeodes, 

 characterised the most primitive Arachnids, then the chelicerEe must 

 be the foremost appendages, and there can be no limbs missing from 

 the Arachnidan head. This view has indeed been recently put 

 forward by Heymons ('01), and adopted by Borner ('02); but it is 

 in dii'ect conflict with the embryological observations of Jaworowski 

 ('91 and '92), who described a pair of evanescent antennae in the 

 embryo of Trochosa, and of Brauer ('95), who describes and figures 

 two ganglionic rudiments in front of the cheliceral ganglia in the 

 embryo Scorpion. That the appendages of one or two segments have 

 been lost from the Arachnid head seems therefore certain ; if so, the 

 mouth of the precui'sors of the class must have been ventral. 



The composition of the head in the Arachnida is of great 

 importance with regard to the subject of the present essay. Doubt 

 has been thrown by later observers on Jaworowski' s observations ; 

 but the appendicular nature of the vestigial antennae that he figm-ed 



