CRUSTACEA PHYLOGEXY 409 



transformation of the stalked eyes into sessile eyes, and disappi'.'innce of the exopo-. 

 dites on the thoracic feet. 



The difference in the position of the heart in the Amphipoda and lanjiodn is to be 

 explained, as already pointed out, by the fact that in the former only a thoracic 

 portion, and in the latter only an abdominal portion of the primitive elongated dorsal 

 vessel has been retained, in both cases, clearly in connection with the localisation of 

 the respiration. The Cumacea stand nearest to the Schizopoda, but exhibit a few 

 Isopodan characteristics. It is pretty certain that the Decapoda come from ScJiizo- 

 j)oda-\\ke ancestors ; the Macrura, and especially the Carididcc, appear to be the 

 most original forms, while the Brachyura and Anomura appear as one-sidedly, but 

 very highly, developed branches of the order. 



We are again met by the question, what is the phylogenetic significance of the 

 various larval forms of the Malacostraca 1 Just as the Nauplius larva, which is 

 characteristic of the whole class of the Crustacea, was held to be a stage of develop- 

 ment repeating the common ancestral form, so a further developed Malacostracan 

 larva, the so-called ZOOM, was considered to be a larva characteristic of the lfc/</- 

 costraca, and assumed to correspond with one of their racial forms. 



These Zocca larvse have been already thus characterised : a large cephalothoracic 

 shield, 2 compound stalked eyes, a median Nauplius eye, head with 5 pairs of 

 limbs, only the most anterior part of the thorax with thoracic feet (2-3) in a 

 rudimentary condition, the remainder of the thorax wanting, or else rudimentary 

 and limbless. Abdomen with full number of segments but without appendages. 

 Tail bifurcated. 



Those investigators who wished to find in this Zocea larva a larval form correspond- 

 ing more or less with the racial form of the Malacostraca had to assume that in the 

 Malacostraca now living the last 5 or 6 thoracic segments are new formations, since 

 they are wanting in the racial form. When therefore the structure and systematic 

 position of the Lepfostraca were better appreciated, and the larval history of the 

 EuphausidM and the Carididce better and more completely understood, these authors 

 were obliged to take refuge in the following explanation. The earliest ancestors of 

 the Malacostraca possessed the full number of trunk segments and the full number 

 of thoracic feet, but a later form lost the last 5 or 6 trunk segments with their appen- 

 dages, while, however, they were still retained by the larva. Finally these segments 

 and their appendages again appeared (in the living Malacostraca known to us or in 

 fossil forms). In support of these very forced views the larval history of certain 

 Stomatopoda was cited ; in these the 3d, 4th, and 5th pairs of thoracic feet are 

 present in the young larva, but disappear later, to be finally reformed. In the larv;i/ 

 of Sergestes also, the last 2 pairs of thoracic feet are reduced and appear again later. 



So many objections, however, can be brought against this assumption that it 

 must necessarily be relinquished. 



In the first place it must be pointed out that of all the manifold known forms of 

 Crustacea, and especially in that series (which indeed is not without gaps) beginning 

 with the Phyllopoda and ending with the Brachyuran Decapoda, of which Bram-lupus, 

 Nebalia, Euphausia, Penaeus are the most important, there is nothing to give the 

 slightest indication that whole regions of the body with their extremities can 

 disappear and reappear again later, or that between already existing segments new 

 segments with their limbs can be intercalated. 



We must therefore ask, to what extent the Zocea larva above characterised is 

 distributed among the Malacostraca. In its typical form it is found only in the 

 Brachyura and (though slightly deviating in form) in the Stomatopoda. This fact 

 alone warns us to be careful, since the Brocln/iirn are decidedly the most recent and 

 most specialised Dcccqmda, and in their development the Schhopodan stage with the 



