26o DISCOVERY REPORTS 



majority of the individuals fall within the lengths shown by the thicker part of the 

 column, very few within those shown by the thin — often one individual only, of the 

 length indicated by the end of the thin part. The number of the Furcilia stage and 

 the number of the individuals measured are inserted in each column. 



The range of development in the individuals falling within one of the later stages is 

 bigger than that in those of an earlier stage. A particularly big range was found in the 

 latest stages of E. triacantha, even though I saw fewer numbers of them than of the 

 corresponding stages of the other two species. In an exceptionally small individual of 

 stage VII the gills were no farther developed than is usual for stage IV (see p. 285). 

 This overlapping in range of size and degree of development is no doubt to be accounted 

 for by the fact that in some individuals the reduction in the number of terminal spines 

 does not follow the path that it follows in the majority : from seven to five, five to three 

 and three to one. From a small sample of E. vallentini with seven terminal spines it was 

 seen that some would have but three in the succeeding moult (see above). It seems im- 

 probable that they would then be as far developed in other ways as three-spined speci- 

 mens that had passed through a five-spined moult. If this is right the Furcilia stages 

 recognized by the number of terminal spines on the telson are not such natural groups 

 as those recognized by the character and number of the pleopods. Fraser (1936, p. 40) 

 expresses the opinion that the Euphausiacea "are arthropods in which continuous de- 

 velopment is giving way to metamorphosis — already well defined in the Nauplius, 

 Metanauplius and Calyptopis stages, less well defined in early Furcilia by the presence 

 of dominant forms and still less intelligible in later development". 



POST-LARVAL DEVELOPMENT 



Fraser's latest Furcilia stage in E. superba, and mine in my five species, is that having 

 one terminal spine and the original three pairs of postero-lateral spines, however modi- 

 fied, on the telson. If we were in accord with earlier workers our latest Furcilia stage 

 would be the same as their latest Cyrtopia stage. In this we are in accord, at least in 

 part, with Sars. Although his definition of the Cyrtopia (1885, p. 150) does not make it 

 clear where that stage ends, his definition of the post-larval stage which follows, does : 

 " All the legs developed. Telson assuming its definitive form and armature."^ He describes 

 the telson of the last Cyrtopia of E. } krohnii (p. 164, sub E. pellucida) as still having three 

 pairs of postero-lateral spines, and figures the telsons of the last Cyrtopia and the first 

 post-larval stage of Thysanopoda tricuspidata (pi. xxxi, figs. 21, 22) as having, respectively, 

 three pairs and two pairs of postero-lateral spines. 



Other workers have not retained this division between the Cyrtopia and the post- 

 larval stages. Lebour (1925, p. 811) wrote: 



Sars regards the post-larval stage as having all the limbs fully formed, and only differing from the 

 adult in the absence or presence of certain setae. The present observations show, however, that the 

 Cyrtopia merges imperceptibly into the adult, and that both male and female may be mature before the 



^ The italics are mine. 



