SYSTEMATIC ACCOUNT 329 



encountered, sometimes in enormous quantities, around South Georgia, the South 

 Sandwich Group and the South Shetlands. It was common in the Peru Current, hut 

 was not observed in the material from the stations in the Brazil Current. This phase, 

 which was represented in the south by large and usually well-developed specimens, was 

 found often mixed with less strongly siliceous, and often smaller specimens which ex- 

 hibited certain "criophilum" phasic influences. These specimens possessed relatively 

 thin walls and the bristles were about the same length as the pervalvar axis of the cell, 

 seldom bearing spines, and making an angle with that axis of about 6o°. The corona of 

 fine hairs was seldom present, and if present, very difficult to see. The connective zone 

 was thin and the scale-like markings frequently were absent or visible only in certain 

 areas on the connective zone. 



The "hystrix" phase is represented in the northern hemisphere by these smaller and 

 less strongly siliceous specimens. It is common in the North Atlantic and has been ob- 

 served in the North Sea, English Channel and the China Seas. It was observed fre- 

 quently in the Southern Ocean, common around the south of Africa, Bouvet Island, the 

 Weddell and Bellingshausen Seas, and in the Drake Strait. 



"inerme" phase. (PI. VIII, fig. 9.) 



This phase is probably seasonal, but the areas in which it was observed are very 

 limited. The cells are robust, usually strongly siliceous, and united to form straight 

 chains. The formation of the chain is rather peculiar, as all the cells are retained within 

 the parent connective zone, and bristles are observed only upon the terminal cells. 

 Frequently the bristles are not free, but held within the connective zone, giving the zone 

 the appearance of being longitudinally striate. In some specimens the bristles are en- 

 tirely absent. Terminal coronas of hairs are seldom present. Owing to this method of 

 chain formation, the valves are only weakly convex. 



Specimens typical of this phase were observed off South Georgia, and again a little 

 below latitude 6o° S, to the north of Enderby Land. 



From the illustrations provided by Karsten, it is clear that he encountered this phase 

 also, and created for it a new specific name (Karsten, 1905, p. 104, pi. 13) which has 

 been adopted to describe this phase. 



"pelagicum" phase. (PL VIII, fig. 2.) 



This phase was encountered but rarely and must be regarded as a warm-water phase. 

 It corresponds to Corethron pelagicum Brun. The chief differences between it and the 

 other phases are considerable shortening of the connective zone, and increased diameter 

 of the cell. This phase is found living free. The diameter of the valve is usually from 

 100-150/x, and the pervalvar axis of the cell is seldom more than 180^. The cells are 

 often quite spherical, the valves being fully rounded as a part of a circle. The bristles 

 are numerous and relatively short. The chromatophores are numerous cocciform bodies, 

 often clustered together. 



This phase was observed in small numbers only around South Africa. It is said to be 



common in the Mediterranean. 



23-2 



