GEOGRAPHICAL DISTRIBUTION. 207 



the exact range of each species, and thus to see how far the limits of the 

 littoral faunae recognized among Mollusca hy Forbes and Woodward, and 

 among Crustacea by Dana, can be accepted for the Echini. Our limitation 

 of a fauna seems as arbitrary as that of a geological period. We take a cer- 

 tain assemblage of animals at any special point, and, contrasting them with 

 a different adjoining assemblage, frequently conclude that because we find 

 differences we are entitled to consider this a different fauna. Such may or 

 may not be the case on a more careful analysis of the range of the elements 

 composing these two assemblages, but too often it will turn out, when carefully 

 analyzed, that what we call a fauna is simply an overlapping of the range 

 of two more distant faunas, giving the intermediate district a peculiar phy- 

 siognomy not always readily separable into its constituent elements, either 

 from want of material or from imperfect identifications. It is frequently 

 almost impossible to determine to which littoral district certain points belong, 

 from the even balance held by the elements of the various districts found 

 there. At the Cape Verde Islands we find a nearly equal number of species 

 known to inhabit the Mediterranean, the West Indies, and tropical Africa ; 

 but there is no reason for making a special fauna of that part of the coast, 

 any more than the mingling of the arctic species of the west coast of 

 Norway with the Lusitanian, British, and even Mediterranean or Atlantic 

 species, entitles the Scandinavian shores to rank as a littoral fauna. 

 Nor does the presence of Arctic, West Indian, and boreal American species 

 south of Cape Cod entitle that region to form a special littoral fauna. It 

 has nothing peculiarly its own, as far as Echini are concerned, and is simply 

 an assemblage of species ranging far beyond the limits which the greater 

 number of species have. It is our ability thus to eliminate foreign elements 

 which will give our divisions greater or less aceuracy. Nor does the presence 

 of Atlantic species along the Virginia and North Carolina coast, associated 

 with West India species, entitle this region to be considered as a 

 special faunistic littoral division. It is a very peculiar littoral district, 

 as much so as the one immediately south of Cape Cod ; but it is an as- 

 semblage of foreign elements which, although they apparently give to the 

 coast of South Carolina, north and south, a very peculiar aspect, so far as its 

 Echini fauna is concerned, is yet made up of species (component elements) 

 having a very wide geographical distribution, and forming parts of totally 

 different combinations. It may be useful to have names for these assem- 

 blages, but they have no biological value. At first glance the presence of 



