20 Mr. H. M. Bernard on 



The shape of the whole fused mass has been no doubt much 

 altered by the action of muscles, but its essential nature as a 

 fusion of metamcricallj recurrent apodemes cannot be 

 mistaken. 



On comparing the segmentations of the two Arachnids we 

 have so far considered, we find that the apodemes forming the 

 endosternite of Oaleodes correspond only with the second 

 pair of apodemes forming the endosternite of Mygale. The 

 first pair in Mygale is now only represented in Galeodes by 

 a small pair of chitinous plates attached by tendons to the 

 cuticle and to the endosternite (fig. 10), and apparently 

 carrying the proximal ends of he main ducts of the coxal 

 glands. 



The endosternite of Phrynus (fig. 6) is a broad crescent- 

 shaped plate. It apparently only has one attachment to tlie 

 ventral surface, and that is to the intersegmental ni^mbrane 

 between the second and third pair of limbs corresponding with 

 the first pair of apodemes fornring the endosternite of Mygale ; 

 dorsally, however, it has four pairs of ligamentous attach- 

 ments to the dorsal wall. These remind us of the endo- 

 sternite of Mygale^ and a comparison of the drawings seems to 

 bear this out. If the endosternite of Phrynus prove uUi- 

 mately to be homologous with that of Mygale^ its sternal 

 attachments, except the first, must have secondarily atrophied. 



The endosternite of Thelyplionus (fig. 7) * is again of 

 special interest, as it clearly follows the same type as that of 

 Mygale and Phrynus. But inasmuch as the fusion of the 

 segments shows no such radial arrangement as in tlie former, 

 nor such longitudinal compression as in Phrynus, the various 

 pairs of apodemes neither radiate from a centre nor form a 

 short crescentic plate, but a long fenestrated framework. 



On turning to Scorpio we find that the endosternite most 

 resembles that of Galeodes. Fig. 8 is from a dissection of 

 Palamncvus Thorellii^ Pocock, made in order to ascertain 

 the exact points of attachment to its parent cuticle. The 

 nerves to the first pair of logs are seen in front of these 

 attachments, which thus correspond with the points of origin 

 of the endosternite in Galeodes. Lankcster describes and 

 figures t a second pair of ventral attachments — " coruua or 

 anterior tendons of the subnoural process ; " but this sub- 

 neural process belongs to the *' diaphragm " and not to the 

 endosternite, as I shall presently show. 



* I have bovrowod this di*awing by Tavunni from Scbimkewit^cb s 

 recent paper iu tbo 'ZouL Aiizeiger' (July IbO.?), "iSiir k Sti-uctiuv et siir 

 la Si^uilicatiou do rendostornite des Aracbnides."' 



t Lankostor, "On tbo Mu.^eular imd liludoskeletal Systems of Limidus 

 uud tScorpio^^ Traus. Zool. fcjoc, vol. \i. part 10 (Icboj. 



