432 PROF. G. O. SARS OX 



the preceding larva ; but at its distal inner corner, in place of the 

 simple conical projection found in that larva, a well-defined 

 cylintlric stem has been formed representing the flagellum. This 

 stem is about the length of the basal part and exhibits a slight 

 trace of subdivision behind the middle. It terminates in a spmi- 

 form process from the base of which a rather long ciliated seta 

 orio-inates. The antennal scale is rather unlike that m the 

 preceding larva, and agrees more in structure with that generally 

 found in Carideau Zoda\ It exceeds the tiagellum by about one 

 quarter of its length and is of narrow oblong form, slightly curved 

 outwards. Its dtstal part is narrowly produced and divided by 

 well-marked transverse sutures into three successive joints 

 o-iudiially diminishing in size. The scale carries inside and at the 

 tip a reo-ular vow of eight ciliated seta? and has, moreover, outside 

 the tip "a very small liair-like bristle. Another similar bristle 

 is also seen originating from the outer distal corner of the 



proximal joint. , ., ., ^- i 



The liiamiihles and maxlUci^ do not seem to exhibit any essential 

 difference in their structure from those in the preceding larva, 

 and are scarcely as yet functionally developed. 



The three pairs of maxillipeds (tigs. 6-8) likewise exhibit mucli 

 the same structure, though the difference in size of the endopodite 

 is still more sharply marked than in the preceding larva, that of 

 the 1st pair (fig. 6) being extremely small as compared with those 

 of tlie two succeeding pairs, the size of which is also somewhat 



unlike. , , , , ^ 



Of the three pairs of imperfectly developed appendages suc- 

 ceedino' the maxiUipeds the two anterior ones (fig. 9), representing 

 the developing chelipeds, are bifid or cleft into two nearly equal 

 dioitiform rami, one of which is the endopodite, the other the 

 exopodite. The posterior pair (fig. 10) are still undivided and 

 shorter than the two anterior ones. They represent the first 

 pair of pereiopoda, the last two pairs being not yet formed. All 

 these appendages exhibit a very simple structure and are filled 

 up with indifferent cells, no articulation or setous armature being 



observable. , . ,■■ 



Ocmrrence.—Oi this larva also several specimens, exactly 

 a-reeing with each other, were found. They occurred m txyo of 

 the sainples, the one taken at Niamkolo, the other at Mbete, 

 both located at the south end of the lake, 



Laeva No. 3. (PL LIX.) 



This larva represents a much more advanced stage than the 

 two preceding ones, and therefore can be determined with more 

 certainty. I think that I am right in considering it as the 

 last larval stage, the so-called Mysis-stage, of Livinocandma 

 parvula Caiman. The solitary specimen which has come 

 under my notice, was found in a sample which contained severa 

 young specimens of that pra.N'n, and both as to size and general 



