262 B- H. BtjxTON, 



these three systems are less developed, possess in addition seven 

 smaller systems of striated tubules in the iirst fourteen segments. 



The similarity of the salivary gland of Perqmtus to the coxal 

 gland of the solifugae is most striking (diagram Fl and 2) and 

 there can be little doubt that they are homologous organs ; the long 

 narrow sac in each instance being an extension of the original 

 modified part of the labyrinth which can still be recognised in the 

 typical nephridia of Peripatus. Both have this extensive secreting 

 sac, which may or may not be filled with products of secretion 

 (photos 37, 37 a of solifugae and 51, 52 of Feripatus), and the outlet 

 on the second appendage close to the mouth opening. 



The saccule in both cases lies anteriorly to the collecting tubule, 

 and the collecting tubule itself passes posteriorly to the labyrinth 

 sac; a point which has not been sufficiently indicated in the diagram (F). 



With regard to the outlet in the solifugae we may infer that 

 as the mouth opening extended forwards the exit tubule of the 

 salivary or coxal gland had to follow it along the coxal joint of the 

 palp, and finally reaching a point where the joint begins to turn 

 outwards (Frontispice B of entire animal, ventral view), the exit tubule 

 itself could extend no further forwards so developed the squirting 

 arrangement which has been described, and the chitinous sheath to 

 keep the fluid in the right direction. 



The coxal glands of the scorpions, pedipalps, and spiders on 

 the other hand appear to have lost the modified part of the laby- 

 rinth and resemble in a general way the large nephridia found in 

 two of the segments of Peripatus, except that the outlets are on the 

 third or fifth appendages with indications of lost outlets on the 

 fourth and sixth, and not on the sixth and seventh as in Feripatus. 



In the scorpions and spiders, as we have seen, special poison and sali- 

 vary glands have been developed for these parts of the work, but such 

 special glands are absent in Tarantula. The outlet of the coxal gland of 

 Tarantula on the third appendage is not far from the ventrally placed 

 mouth, and we have seen that in the middle part of the labyrinth the cells 

 are columnar and basophil, so it is possible that they act as secreting 

 cells, and that the coxal gland in this case also plays the part of 

 salivary as well as of excreting gland. However these cells of the 

 middle part of the Tarantula labyrinth have the usual striated base 

 on which the nuclei rest, and do not resemble in this respect the 

 secretory cells of the labyrinth sac in Feripatus and Galeodes. U 

 then these cells are actually secretory it is probable that they have 



