298 ME. GEOiiGE M. THOMSON OX A EEESHWATEK 



by the presence of a well-marked lateral suture on the tirst thoracic segment, recalling 

 the epimeron of most Isopoda, which, however, is in some instances not developed on 

 the first, though prominent on the succeeding segments. 



The plate-like character of the branchiae is a feature which does not exhibit nearly so 

 radical a digression as the loss of the carapace. In the development of Ewphaiisia 

 pellucida, Sars has shown * that after the Cahjptopis stage the larva passes through a 

 Furcilia stage, in which the anterior legs and the pleopoda begin to develop. The legs 

 first ajipear as simple processes or lobes, giving rise on their outer side to a minute knob, 

 which is the rudimentary exopodite. As the endopodite develops and gradually becomes 

 articulated, a simple; knob-like protuberance arises at its base, below the exopodite, 

 which as development proceeds bifurcates into two rudimentary plates. This stage 

 appears to cori'cspond with tliat reached by Anaspkles, only in the Furcilia larva the 

 branchiae continue to subdivide imtil they ultimately become arborescent in the adult, 

 because they are sheltered under a carapace and so must expose the largest surface 

 possible to the somewhat limited sujjply of water with which they are surrounded. In 

 most of the Schizojioda the gills become arborescent. 



The archaic character of Anaspides is more fully revealed by a study of its anatomy. 

 The alimentary canal sho\\s a much simpler structure than prevails in any other 

 described Schizopod. In Ilysis t, which may l)e looked on as one of the least highly 

 developed forms of the suborder, the stomach is distinctly divided into an anterior 

 (cardiac) and a posterior (pyloric) })ortion ; the latter is much the smaller, is of very 

 complicated structure, and is the j^art in which the food is chietiy triturated. Opening- 

 just behind it and placed dorsally is a small caecum, which acts probably as a sort of 

 salivary gland. The liver is composed of ten jjouch-likc masses, five on each side, which 

 unite their contents into a hepatic duct entering the intestiiie just where it leaves the 

 stomach. The intestine is a nearly straiglit tube, passing directly to the anus, and having 

 no caeca or diverticula communicating with it anywhere behind the opening of the 

 hepatic duct. In Ettphausia % the structure of the alimentary canal is essentially similar, 

 but above the pyloric portion of the stomach are two small incurving caeca (not one as 

 in Ml/sis), while the liver consists of a mass of minute hepatic tubes on each side of the 

 intestine and not reaching to the posterior portion of the thorax. These masses of tubes 

 are quite distinct in appearance and aggregation from the long isolated tubes of 

 Anaspides. Ca^ca of the mid-gut are found in all orders of Crustacea as paired or 

 unpaired organs. Their position appears to vary a good deal. Eor example, in some 

 Bracliyura (e. g. Carcimis) a pyloric ca3cuni or diverticulum arises on each side of the 

 stomach in front of the hepatic ducts, which when uncoiled exceeds the intestine in 

 length, Avhile at about one-third of the distance from the stomach to the anus a solitary 

 caecal appendage arises, also in the form of a long tube. In others (e. g. Maia) the 

 latter caecum is placed at the extreme anterior extremity of the intestine. In the 

 Macroura there is always the usual pyloric caecum, but in some genera (e. g. Astacus) 



* L.c. p. 10:?, pi. XXX. figs. 23-27. 



t G. 0. Sars, ' Hist. Nat. des Urustaces d"eau douce de Norvi'go/ p. 2G. 



X ' Challenger ' Report, p. T-j. 



