NORTH ATLANTIC SEA AND THE FAROE CHANNEL. 129 



occurrence. What was meant by E. 2}iilehra (frequent at 450 to 320 

 fathoms) in Fowler's list {P.Z.S., June 21st, 1898) I do not know, but 

 I am pretty confident that it was not E. imklira, and probably was 

 E. rostmta. I found the latter species in Fowler's collection, and I 

 have taken it several times since, but it has not occurred in my collec- 

 tions in the Atlantic south of the Wyville-Thompson ridge, being there 

 replaced by a variety of E. eurticauda. E. rostrata does not apparently 

 occur in the Norwegian Sea, but it is recorded by Scott in the Ceylon 

 Copepoda. In the neighbouring Maldive Islands, however, I have not 

 met with it, but only E. hella. That it should occur at such widely 

 different localities is not a little curious. The limitation previously 

 given by Giesbrecht {F. v. Fl. N.) was 44° N. to 41° S. 



Paracalamis parvus. Found plentifully in Christiana Fjord by Sars, 

 and South Norway, though not apparently further north, none having 

 been observed at Bergen by Nordgaard, nor in the Plankton samples 

 from the Northern Ocean examined by G. 0. Sars (Sars' Crust. Nonvay, 

 p. 18), and common round the British coasts. It occurs in the Faroe 

 Channel, and as far south, at any rate, as lat. 51^ (Valentia in Ireland) ; 

 occasional in deep-water samples, it is not common in the open ocean. 

 It has probably not such a wide southern distribution as has been 

 imagined. Met with in the Indian Ocean and Mediterranean, I think 

 there is reason to differentiate the two forms, boreal and Mediterranean 

 (and Indian Ocean), as at least distinct varieties. A careful examina- 

 tion of the figures of this species given by Sars (Crust. Norway) and 

 Giesbrecht {F. u. Fl. Golfes Neapel) discloses differences, and I have 

 made a detailed examination of examples from the Faroe Channel and 

 from the Indian Ocean (Maldives) for the purpose of comparison. 



The Southern Ocean examples are found constantly to be rather 

 smaller than the northern, the basal joints of the feet are more 

 densely armed with short spines (in the northern variety these are 

 almost entirely absent, especially on the fourth pair), the basal joints 

 and the last segment of the exopodites (fourth pair) are broader in 

 proportion to the length (exopodite 3 is five times as long as broad in 

 the boreal variety, only four as long as broad in the Indian variety), 

 and the anterior antennai are rather longer in proportion to the body 

 in the southern variety. On the whole the southern variety may be 

 said to be constantly smaller, more spiny, and with less attenuated 

 segments of the feet (in which the marginal teeth are also stronger 

 and coarser) than the northern. This difference is also noted in 

 comparing the figures of Giesbrecht's P. parvus from the Mediter- 

 ranean with those of Sars' P. loarvus from Norway. They are not 

 distinct species, but undoubted varieties, and the northern form, 

 though extending as far south as lat. 51° (Valentia), does not prob- 



NEW SERIES.— VOL. VII. NO. 1. I 



