X18 DR. J. D. F. GILCHRIST ON LARVAL AND 



the laro-est spines of the adult. No other spines were seen on the carapace. 

 There are no sternal spines. 



The rostrum is a very small simple projection, and does not appear to meet 

 the ocular segment. 



The antennules (PL 13. fig. 33, a. 1) are relatively short, the peduncle 

 reachin"- to about the first third of the fourth segment of the antennae. 

 The flagella are, however, relatively stouter and longer than in the adult. 



The antennce (fig. 33, a. 2) are much longer than in the largest phyllosoma. 

 The peduncle has the chief large spines of the adult, but the smaller scale-like 

 spines are not yet developed. 



The mandibles (PI. 13. fig. 34), of all the other parts, show the greatest and 

 most abrupt change from the phyllosoma conditions, but have not as yet 

 assumed the adult characteristics. They thus present an instructive inter- 

 mediate stage. The change is doubtless associated with the assumption of a 

 ground habit, the mouth-parts being used for crushing up the harder shells, 

 etc. on which the animal now feeds. The setse, so prominent in the 

 phyllosoma, are replaced by a thick cuticle. The incisor part can still be 

 distinguished from the molar ; it consists of a broad cuticle with two slight 

 projections, which in some are hardly distinguishable, but are clearly repre- 

 sented in the subcuticular tissue, and, in later stages, become the two 

 prominent teeth of the cutting-edge of the adult mandible. The molar part 

 is in the form of a blunt well-marked projection, which has now come to lie 

 somewhat behind the cutting-edge. In other words, the free edge of the 

 mandible of the phyllosomas, consisting of cutting and mandibular parts, 

 instead of forming a slightly curved edge, is now bent so as to form almost a 

 circle. This can be clearly seen when the mandible is viewed from its distal 

 extremity. It would appear therefore that the molar part of the mandible of 

 the phyllosoma in the course of its development turns backward behind the 

 cuttino-edge, and assumes the flat hard character of the molar part of the 

 adult mandible. The endopodite of the mandible, no trace of which was 

 found in the phyllosomas, now appears, but consists only of one distinct 

 seo-ment. The beginning of the mandibular spine is also seen. 



HhQ first maxilla (PL 13. fig. 35) has now assumed the foliaceous form of 

 the adult. The protopodite is deeply cleft into two segments, and near the 

 base of the second a small endopodite appears, about half of its base arising 

 from the inner side, so that it is not clearly seen when viewed from below. 

 All the spines have disappeared, though a few were seen on the anterior 

 border of the coxopodite and basipodite on one specimen. The appendage is 

 now removed from the base of the lower lip by a distance equal to the length 

 of the coxopodite. 



The second maxilla (PL 13. fig. 36) has the three endites of the protopodite 

 relatively larger than in the adult condition and provided with a few setse. 

 The endopodite is relatively much smaller and is triangular instead of convex, 



