1855.] Linnean Society. 377 



carapace in Brachynra, half the same in Macroura, and lessens in im- 

 portance as the animal descends in the scale of nervous develop- 

 ment. 



In Diastylis and Cuma, the eye (for in this family the two 

 coalesce and form but one) is nearly in the centre of the carapace ; 

 this position is the result of the great development of the lateral 

 angles of the posterior portion of the carapace, which meet in front 

 and form what appears to be a rostrum, yet never unite, but continue 

 distinctly separate, through the median line of the so-called rostrum, 

 which separation is persistent on each side as far back as the 

 posterior portion of the third ring, and then continues in the line of 

 fusion across the back where the two meet*, which line homologizes 

 with that which Milne-Edwards has named the cervical suture in 

 the Macroura. 



The constant relative position of this suture in all Crustacea must 

 be the same ; it forms a line of demarcation between the third and 

 fourth (that is the posterior antennal and the mandibular) rings. 

 The posterior antenna, anchylosed as it is with the dermal skeleton 

 in all the Brachyura, still holds the same relative position as in the 

 pupa stage ; therefore by inversion, since the ring folds over so as to 

 form the orbit, the anterior limit of the cervical suture must be 

 beneath and on the inner side of the posterior antenna. Such a 

 suture is plainly demonstrable in most of the Brachyura, but it 

 extends posteriorly to the extreme limits of the carapace, forming as 

 it were two wings or side pieces, the epimera of Milne-Edwards f ; 

 this line, according to the author's opinion, homologizes with the 

 cervical suture in the Macroura, the portion anterior to which is 

 most developed when centralization of the nervous system is most 

 perfect, and vice versd. Centralization decreases as the posterior 

 portion of the carapace increases. In the Brachyura, this line of 

 union is more or less perfectly fused, but sometimes splits when the 

 animal throws off the exuviee. Here the mandibular ring or portion 

 posterior to the cervical suture is at its minimum, and the antennal 

 ring or anterior portion is at its maximum. 



In Macroura, where the nervous centre commences its first ten- 

 dency to separate into distinct ganglia, we find the anterior portion 



• The figures of Cuma Hathkei, in Kroyev's great work on Scandinavia, 

 agree with those given with this paper ; a circumstance, of which the author 

 was not aware until be had perfected his own researches. 



f That the epimera of Milne-Edwards homologize with the mandibidar 

 ring, has been previously stated by Dana, a fact of which the author was not 

 aware until the day on which the paper was communicated to the Society. 



