482 DEPARTMEXr OF THE XATAL SERVICE 



todentata is a Gulf Stream form, but its occurrence in the boreal water is not indi- 

 cative of a recent Gulf Stream origin, since it is of a decidedly different type in the 

 boreal water. Its abundance in our boreal water and its rarity on the European coast 

 may be due to its ability to live and grow to maturity but not to reproduce success- 

 fully in boreal water. With this interpretation our boreal water would have a very 

 slight but constant contribution from the Gulf Stream. 



Of tlie deep-liviijg species, 8. hexaptera is absent from the boreal wa<ter, but 

 *S'. lyra was found at two of the boreal stations. One of these stations was, however, 

 really on the edge of the Gulf Stream, and the single specimen found at the other 

 station (Acadia station 70) was much larger than any others obtained. This indi- 

 vidual may have passed through the bottom of the Gulf Stream, since the species goes 

 into very deep water and is a constant inhabitant of the depths of the Atlantic ; or if 

 it has entered the boreal water by the mixture of the latter with the Gulf Stream, its 

 size precludes a recent entrance. 



We have therefore no certain evidence of any deep contribution of Gulf Stream 

 water to the boreal zone in our region, and evidence of only a slight surface contri- 

 bution. 



Of movement in the opposite direction, from the boreal water to the Gulf Stream, 

 since there is no peculiar surface boreal form, we have merely the negative evidence 

 of rarity of the surface tropical species at the northern Gulf Stream stations. The 

 deep-living boreal species, <S'. maxima, was regularly found at the Gulf Stream stations 

 except at the extreme southwest, and it was more abundant at the north, while 

 Eukroknia liamata was found at only one of the stations, the most northerly {Acadia 

 station 74). There is therefore evidence that the boreal water does contribute to the 

 Gulf Stream in the deeper part, and perhaps also at the surface. The latter contri- 

 bution will tend to be indistinguishable from coastal water. 



In the Northern Oceanic zone we have large ,S^. serratodeniata at the surface, and 

 in the depth, 8. maxima and E. hamata. For its extent in July- August see figs. 3, 

 6 (vertical lines), and 12. There is to be seen an extension of the surface water over 

 the Scotian bank to the south, and of the deep water up the Laurentian channel to the 

 7iorth. The deep water is present in small amount over the Scotian bank, the surface 

 .vater in the Laurentian channel, and both in the gulf of Maine and Bay of Fundy. 

 The virtual absence of this water over the Newfoundland banks is worthy of note, it 

 being held off by the coastal water. Its centre is seen to be a narrow zone close against 

 the continental shelf, decreasing in width to the southwest. The vertical distribution 

 of 8. maxima and E. liamata shows that it passes to a deeper level below the coastal 

 water up the Laurentian channel and below the Gulf Stream to the south and south- 

 west. 



On the May-June cruise the small area explored showed a similar distribution, 

 but extending farther out from the continental shelf. 



The absence of any continuation of the deeper part of this water (as indicated by 

 E. liamata and 8. maxima) along the continental shelf oft" Shelburne in July- August 

 as shown by our most southern section, indicates the abruptness of the transition from 

 Gulf Stream to coastal water at this point, the boreal water having been squeezed out 

 by the pressing in of the Gulf Stream close to the continent. This is doubtless 

 temporary. At another time, perhaps earlier in the season, the boreal water would 

 be much more extensive and pass continuously down the coast and up into the gulf of 

 ]\raine. 



Dawson has shown that in the deep parts of the St. Lawrence gulf there is water 

 with the characters which we have given above for boreal oceanic. The Albatross, as 

 reported by Townshend, found in July, 1885, very low temperatures at the bottom 

 which would indicate no boreal water on the bottom on the banks just south of New- 

 foundland, on the Breton bank, nor close along the shore of Nova Scotia; but in the 

 mouth of the Laurentian channel and off Shelburne near La Llave bank the bottom 



