141) 



TAOK SKOr.SUEUr, 



lu the sample that was investigated there was, as is shown above, no fi-ce-living hirvai 

 class younger than that of 29d. But the hitter is probably not to be taken as the youngest 

 free-living stage. In a number of females larvae were observed which were apparently ready to 

 leave the brood chamber. These larvae were only 0,9 mm. (= 20 d) long. If, using Brooks's 

 law, we di\-ide 29 d by 1,23 and the quotient again by 1,23, we shall obtain, of course, the 



theoretical average lengths of the two next 

 youngest stages. According to this method 

 of calculation these stages ought to have an 

 average length of 24 d and 20 d respectively. 

 In other words the younger of these two 

 stages ought theoretically to have the same 

 average length as that which was determined 

 empirically for those larvae that are about 

 to leave the brood-chamber of the mother. 

 This stage ought consequently to be presum- 

 ably taken as the seventh; on the other hand 

 the larval stage representing the sixth was 

 quite absent in this sample. 



Corresponding classes of length were also 

 found in other samples, but in these the abso- 

 lute measurements were different with different 

 external conditions. In all these samples, 

 however, the younger larval stages were repre- 

 sented still more sparsely than in the one dealt 

 with above. 



Although, as in Cypridina (Doloria) 

 pectinata and Cypridina (Vargula) norvegica, 

 the classes of length that were observed were, 

 at least in a number of cases, very near each 

 other, it seems to me certain that each re- 

 presents a moult; the agreement with the post- 

 embryonal development of the preceding species 

 is too great to leave any doubt. 



The agreement between the develop- 

 ment of the larvae in Philomedes globosa and 

 Cypridina (Doloria) pectinata is, as is shown 

 above, ver}' great, ^^^len to this it is added that quite a similar post-embryonal develop- 

 ment is observed in Cypridina (Vargula) norvegica, it ought not to be too bold to 

 assmne that the post-embryonal developed is, on the whole, quite similar in all repre- 

 sentatives of the family Cypridinidae (perhaps in all the species belonging to the sub-order 

 of Cypridiniformes). 



Fig. XXI. — \ Curve showing the results of my measure- 

 ments of the lengths of the [shells [of a number of 

 specimens of Philomedes globosa from Greenland. The 

 abscissa represents the lengths of the shells expressed 

 in di\isions of a micrometer (29 divisions = 1 mm.); the 

 ordinate represents the number of specimens measured. 

 1) = 1 o 2) = 1 9. 3) = 2 2. 4) = 1 (?. 3) = 1 $. 



