EEPORT ON THE CEPHALOPODA. 211 



I have collected whatever information the specimens in my hands and the litera- 

 ture of the subject afforded in the subjoined lists. 



As regards their distribution the Cephalopoda seem to be divisible into three principal 

 groups (1) the Pelagic, (2) the Littoral, (3) the Abyssal, and the different genera belong 

 with considerable constancy to one group or the other, that is to say, all the species of 

 any one genus belong to the same group; for example— Oci;oj^^/s, Sepia, and Loligo 

 are typical littoral genera, while Cranchia, Ommastrephes, and Onychoteuthis are 

 with equal distinctness pelagic in their habits. 



It has been stated generally, and the examination of the Challenger collection has 

 certamly borne out the proposition, that while pelagic animals belong to but few types, 

 each of which has a comparatively wide area of distribution, littoral forms belong to many 

 species, each of which is confined within narrow limits. With reference to deep-sea forms 

 our knowledge is only in its infancy, but they seem to be even more widely distributed 

 than the pelagic ones ; and conditions of life in the depths of the sea (especially tempera- 

 ture) are so uniform that this is precisely what we should expect, and what has been 

 found to obtain in other groups. 



The first three lists give what I may call the "Oceanic" species, using this word 

 to include both the pelagic and abyssal forms, for it is convenient to consider them 

 together as regards their horizontal distribution: their vertical distribution will be 

 treated of in the next section of this Eeport. These forms have been disposed in three 

 groups corresponding to the Atlantic, Pacific, and Indian (including the Southern) 

 Oceans rather for convenience than from a belief that such a division is natural; 

 although, as will appear subsequently (p. 222), the great majority are confined to one 

 area. The chief factor limiting their dispersion being probably temperature, though 

 doubtless other conditions, such as the presence of Gulf weed, also have their influence. 



The " Littoral " species, that is those found in moderately shallow water not far from 

 the coasts, whether they be active swimmers like Loligo, or more sedentary like Octopus, 

 are much more restricted in their range than the oceanic. For the purpose of represent- 

 ing their distribution, the coasts of the world have been divided into seventeen regions, 

 which are very different in extent and in the number of species that have been recorded 

 from them ; as regards the former of these points it may be remarked that no sharp 

 boundaries can be drawn between them at all ; for, although for statistical purposes it 

 may be necessary to adopt lines of demarcation, these are not recognised by nature, and 

 furthermore a fuller knowledge of the faunas of the various regions would almost 

 certainly show that some of the districts here proposed should be subdivided and others 

 united. 



The geographical regions here adopted agree very closely with those proposed by 

 Dr. Paul Fischer in his recent Manual, based upon a study of the whole of the 

 MoUusca. In a few cases I have subdivided his districts, in more he has subdivided 



