COMPAEATIVE MORPHOLOGY OF THE GALEODID^. 361 



undoubtedly retained a primitive feature which has been lost in Galeodes. Before connecting the two 

 forms, we will describe the coTiditioii in the Thelyphonidie and Pseudoscorpions. 



Thelyphonidce. — As Laurie has described (47), there are four pairs of diverticula from the central canal, 

 alternating regularly with the dorso-ventral muscles. The fourth pair stretches back on each side inside 

 the rows of dorso-ventral muscles, and is itself constricted by them, while the central canal becomes 

 transformed into what is functionally a hind-gut (PL XXXIV. fig. 2). 



Pseudoscorpions. — The canal gives off one median ventral and two lateral diverticula in the first 

 segment ; the latter run backward on each side of the central canal and are regularly constricted by 

 the dorso-ventral muscles (PI. XXXIV. fig. 3). 



I have placed these four types of the alimentaiy canal side by side in PI. XXXIV. : Scorpio fig. 1, 

 Thehjphonus 2, Galeodes 4, and Chernes 3. 



From the segmental character of the diverticula separated from each other by dorso-ventral muscles, 

 we are, I think, justified in assuming that in the primitive form there were originally at least seven 

 pairs in the abdomen before reaching the chitin-lincd hind-gut, which appears to have begun between 

 the 7th and 8th segments. We should then have the primitive arrangement, consisting of seven pairs of 

 diverticula separated from one another by six pairs of dorso-ventral muscles. I have shown this hypo- 

 thetical primitive condition in fig. 5 and also, slightly altered, in PI. XXXIII. fig. 12. 



To what are we to attribute these various modifications of the primitive segmental arrangement ? It 

 is at once apparent that in all cases the posterior diverticula degenerate, while the few anterior diverticula 

 which persist are of increased size and importance. We shall not be far wrong if we see in them 

 various adaptations to the method of feeding. I would throw out the following suggestion. 



The forcing of the liquid food into the alimentary system has to be regulated, otherwise it would be 

 driven straight out through the hind-gut and anus. W^e accordingly find the central canal so narrowed 

 in its course that masses of fseces could easily be utilized to stop the passage. Thus the food would 

 tend to be pumped chiefly into the anterior diverticula, which would become highly developed, while the 

 posterior would degenerate. The longer the liind-gut is, the further back would such stoppage occur, and 

 the greater the number of the diverticula which could develop. Hence the great number in Scorpio. 

 In Galeodes, where the dorso-ventral muscles, meeting ventrally in the middle line, are near together, the 

 greatly enlarged antci'ior diverticula grew backward outside them ; but in Thelyphonus and Chernes these 

 muscles are wide enough apart to allow the anterior diverticula to grow backward between them 

 alongside of the central canal. 



Phrynus is said to have four pairs of abdominal diverticula. I can find only three pairs ; the arrange- 

 ment in other respects closely resembles that of Thelyphonus, the smaller number of diverticula being 

 perhaps connected with the shortening of the hind-gut by the loss of the tail. 



There seems to be great difference of opinion as to the number of abdominal diverticula in the 

 Spiders. Wasmann (74 a) figures two pairs in Myyale. Bertkau, who studied this subject with 

 great care, concluded that, in addition to certain small diverticula, entering irregularly into the central 

 canal immediately behind the waist, there are five primary diverticula — a median ventral, which may 

 compare with the median ventral diverticulum in the Chernetidse, and two pairs of lateral. These 

 median ventral diverticula may perhaps be treated as accessory outgrowths of the mid-gut due to the 

 action of the force-pump arrangement of the oesophagus *. 



In the Phalangidse and Acari we have altogether different conditions. In the former we have a 

 reduction to six abdominal segments, and in the Acari very variable reduction — in some cases the 

 abdomen consists of only three segments t- We should therefore hardly expect abdomiual diverticula 

 to be developed in such cases. 



* Bertkau (16) caUs these diverticula " glands," under the influence of the old name, the " liver." 

 t Ixodes, which has carried the distensibility of the abdomen to an extreme, develops 10 segments, which, as in 

 the Spiders, early become obscured : Wagner (73). 



SECOND SERIES. — ZOOLOGY, VOL. VI. 48 



