358 ME. H. M. BERNAED ON THE 



above described. This bend in tbe oesopbagais may perhaps be an indication of the 

 compression of the first three segments. 



The mid-gvit itself pursues a straight course backward to the commencement of the 

 hind-gut. It is lined, apparently throughout, by an epithelium of cylindrical cells, which 

 vary in depth in different parts (PI. XXXIII. figs. 10, 11) ; they are specially long just 

 before reaching the hind-gut. 



In nearing the diaphragm, the canal narrows considerably, and in passing through it, 

 its lumen is very small (PI. XXXIII. fig. 1, and also figs. 2, 3, 4, Avhich are drawn to scale). 

 After passing the diaphragm, the lumen increases greatly in size, and then gradually 

 narrows again in the 4th abdominal segment, remaining narrow until it joins the hind- 

 gut (PI. XXXIII. figs. 1 & 5). 



The mid-gut of Galeodes, as of all Arachnids, is supplied with diverticula for the 

 reception of the liquid food. These originally segmental diverticula were, in their 

 simplest condition, probably inherited structures. It is not impossible that there is some 

 relation between their present enormous development in the abdomen, where they fill up 

 every available space, and the pumping-apparatus — that is, the latter may have helped to 

 sjiecialize the former by its force-pump action. 



Por convenience, we divide these diverticula into two groups, the cephalothoracic and 

 the abdominal, although, morphologically, they belong to one and the same series. But, 

 while the diverticula in the muscular cephalothorax show signs of degeneration, those in 

 the abdomen, which is specialized into a distensible sac, are, as stated, developed to an 

 astonishing extent {cf. PI. XXXII. fig. 18 with PI. XXXIII. fig. 5). 



The Cephalothoracic Diverticula. — In the cephalothorax, the mid-gut of Galeodes gives 

 off four pairs of thin diverticula towards the four pairs of legs : the two anterior pairs do not 

 run into the legs themselves, and are apparently atrophying ; they are often found folded 

 back upon the gut (PL XXXII. fig. 18). The two posterior pairs penetrate into the limbs 

 as far as into the trochanter. These two usually have a branch about halfway down their 

 lengths. This tendency to branch in the posterior cephalothoracic diverticula is interesting 

 as leading on to the highly-branched diverticula in the abdomen. The anterior segments 

 of the body were here again clearly those first specialized. These cephalothoracic 

 diverticula are provided with powerful circular (cm) and longitudinal fibres (PI. XXXII. 

 fig. 19, Im). 



These four pairs of diverticula, running towards or into the four legs, seem very typical of Arachnids, 

 being found in Galeodes, the Spiders, Phrynus, Thehjphonits (Laurie), and Phcdanyium (Blanchard), 

 and lead us to conclude that in the primitive form the CES0j)hagus ran through two segments, while in 

 the 3rd segment the mid-gut commenced, sending out lateral diverticula in this and in the following 

 three segments, omitting for the moment all mention of tlie abdominal diverticula. 



That these diverticula were originally separated one from another by dorso-ventral muscles we have 

 sufficient evidence. Although, in Galeodes, these muscles in the cephalothorax are no longer, with 

 certainty, recognizable {cf. supra, p. 336), in the Spiders we have the dorsal suspensors of the endo- 

 sternite regularly separating the diverticula ; the same is apparently the case in Phrynus and Thehjjihonus. 



In those Arachnids in which there are no longer the four typical cephalothoracic diverticula, as 

 Scorpio and Chernes, this is due to secondary reduction. In the former case [Scorpio) it must be 

 accounted for by the extraordinary longitudinal compression of the cephalothorax. Instead of the 



