568 TAGE SKOGSBBRG 



Divi^io/i of Ibis 

 fn III ill/. 



1,0 nun. liiiig. Tlu' m;iture .specimens were assumed to belong to , .Stage 1", the three hirvae 

 to the Uist hirval stage, ,, Stage III". The reason why the mature specimens were assumed to 

 belong to ,, Stage I" was obviously the relatively great difference with regard to the length of 

 the shells between these specimens and ,, Stage III"; these two stages were really presumably 

 separated by an intermediate stage. The question then becomes: did the three larvae really 

 belong to the last larval stage? Unfortunately the statements given are too incomplete for me 

 to venture to say anything quite definite in this matter. It seems to me, however, from pi. XIX, 

 fig. 77, very probable that these three specimens belonged to the next to the last larval stage. 

 If this is the case, then the reason for assuming a mature stage between the stages found by 

 (i. H. Fowler also disappears. On the other hand G. H. FovvLEli found in this material two 

 stages of mature females, ,, Stage I" being represented by seventeen s^jecimens, ,, Stage 11" 

 by only three. Were both these stages mature? For the same reasons as in the case of the males 

 it is very difficult for me to make any statement on this point, but it seems tx) me practically 

 quite certain that the three specimens of ,, Stage 11" were not mature; pi. XIX, fig. 80 definitely 

 shows this. They were probably larvae in the last stage. If this is the case, there was in this 

 sex too only one mature stage. 



What has been said above will be sufficient to show clearly how uncertain is the basis 

 on which G. H. Fowlek has constructed his important hypothesis. 



As is seen above, p. 564, this group was divided by (1. W. Muller, 1906 a, into two sub- 

 families: Thaumatocyprinae and Conchoecinae. The same classification is also used in the 

 present treatise. Of these two sub-families Thaumatocyprinae , which is so interesting from 

 a systematic point of view, was unfortunately, however, quite unrepresented in the collections 

 investigated by me. 



Sub-Family Conchoecinae. 



Sul)-Fam. Conchoecinae, (J. W. MOLLi:!:, 1906 a, p. 43. 



Description: — Shell: — This is dimorphous, but in a numl)er of cases only rather slightly 

 so. — Tile rostral incisur is shallow in all species, but it never seems ti i be (juite al)sent. An apparent 

 deepening of the incisur occurs, however, in all tlic forms so far known. This deepening has arisen 

 because the outer lamella of the shell curved out like a pocket just above the incisur, forming a 

 rostrum which is in most cases rather extensive (this rostrum is thus not homologous with the part 

 with the same name in the Cypridiniformes); the original anterior margin of the shell continues 

 (as 0. W. MtiLLEPv pointed out as early as 1894, p. lol) in the shape of a more or less S-shaped 

 curved line (,,Buchtlinie", according to ( '. ('l.ALiS's terminology) proxinially on the inside, or 

 perlia])s more correctly speaking, on the ventral side of the rostrum. The rostral incisur is always 

 situated al)o\'(' half tli(> heiglit of the slirll. in must cases (|uite near its dorsal margin 



