438 PROF. G. O. SAES ON 



most one occupying the outer distal corner, the other two placed 

 at some distance from each other and more dorsally. 



The uTopoda only differ from those of the preceding stage in a 

 somewhat richer supply of marginal setae. 



Occurrence. — The above-described specimen was found in the 

 same sample as the preceding larva, and was associated with three 

 other somewhat larger specimens. Moreover, young specimens of 

 the same species occurred in two other samples, the one taken at 

 Mbete (S. end of lake), the other at Karema (E. shore of lake). 

 In the latter sample numerous specimens were present. 



General Remarks. 



Among the Macrura inhabiting fresh water, so far as I know, 

 the genus Caridina is the only one in which a true metamorphosis 

 has been found to occur. In the other forms the postembryonal 

 development is either direct, as in Astacus, or at least exhibits 

 only very slight traces of a metamorphosis. This difference in 

 the development, according to the medium, may even be found 

 in a single species, viz., in the case of Paloemontes varians. The 

 typical form of this species, found in salt or brackish water, 

 passes through a quite normal metamorphosis ; whereas in the 

 variety inhabiting purely fresh water the larval development is 

 much abbreviated, the young leaving the egg with all the append- 

 ages present, except the uropoda. A similar suppression of the 

 metamorphosis is certainly also observed in some of the marine 

 forms ; but these are, as a rule, only such as inhabit very deep 

 water. In shallow-water forms, on the other hand, a compli- 

 cated metamorphosis is always found to prevail. In most cases 

 the relative size of the ova, borne by the female beneath the 

 metasome, will be suggestive as to the decision whether the 

 metamorphosis in the observed form is complete or suppressed. 

 If the ova are very large and few in numbei-, this is an infallible 

 sign that the metamorphosis is incomplete or quite absent. 

 On the other hand, the presence of very small and numei-ous 

 ova points to the assumption that the young hatch in an 

 imperfect larval condition, and of course must pass through a 

 more or less complicated metamorphosis before reaching maturity. 

 In most of the prawns inhabiting Lake Tanganyika, as established 

 by the measurements given by Dr. Caiman, the ova are of un- 

 usually small size, and from this fact alone it could be inferred 

 that a true larval development must be present. This has indeed 

 been fully confirmed by the investigations here published, and in 

 one case it has, moreover, been proved that the larvte, on escaping 

 from the ova, exhibit a still simpler and more primitive structure 

 than found in any other known form of this group. 



A very peculiar feature, exhibited by the two Zoese described in 

 this paper, remains to be mentioned, viz., the very rudimentary 

 and as it were embryonic appearance of the oral parts (mandibles 

 and maxilloe). In all free marine larvtc belonging to the present 



