650 MR JAMES MURRAY ON 



species by the spines is mechanically easy, and has been accepted as sufficient by most 

 authorities. Professor Richters requires the animal described to be mature, unless 

 there are some striking peculiarities. 



There is no doubt that the processes are remarkably constant in some species, even 

 among those of complex and unusual armature, such as E. oihonnse and E. islandicus. 

 It is also known that some species, while still in the two-clawed larval stage, possess 

 all the processes which characterise the adult. 



Nevertheless there is reason to believe that the processes vary in number ; and if 

 this can be demonstrated, it will be necessary to broaden the diagnoses in accordance 

 with the facts. If this caution is disregarded, species may be made almost ad lib., 

 with a repetition of the disastrous fate which has overtaken some other groups of animals, 

 or rather the students of them. 



The difficulty is greatest in dealing with the central group of species of simplest 

 structure, all built on the same plan, with two pairs of plates, two median plates (or 

 three, the third being rather uncertain), and V and VI joined into a single 3-lobed plate. 

 The distinctions among species of this group rest on the variations of lateral processes 

 (maximum number 4, excluding the head) and two dorsal processes, on the paired 

 plates, — all of these processes, in their greatest development, being long whip-like 

 setae. If we suppose the suppression of one or more pairs of processes (a common 

 accident of development), and some of the others shortened in various degrees, a 

 great number of changes can be rung on this simple arrangement. Distinctions of 

 texture, which might help matters, have been overlooked, and, moreover, many species 

 of this group have never been seen mature. 



Variation in the number of spines has been noticed in several species. When a 

 pair of spines is suppressed it is difficult to prove variability, but when in a plentiful 

 collection of animals, alike in general build and texture, sometimes one spine of a pair 

 and sometimes both are absent, the presumption is that all belong to one species, and 

 that the spines vary in number. E. spitsbergensis has a very distinct characteristic 

 texture of the plates (though the nature of the markings, which are not papillae, is not 

 known). Many examples have the second dorsal process (a broad triangle as figured 

 by Scourfield) more or less elongate, and a series may be formed showing all gradations 

 to a long seta, like the first dorsal process. Some examples are found which lack seta 

 b, without other difference. 



It is in a form which I here unite doubtfully with Richters' E. quadrispinosus, as 

 var. cribrosus, that the widest series of variations have been traced. This also has a 

 characteristic texture, of unequal dots, which look like perforations, but may be of 

 another nature, and plates II and V + VI are further crossed by plain bands, both 

 transverse and longitudinal. This has typically all four lateral processes, after the 

 head, as long setae, and the dorsal processes as short curved spines. One variation 

 shows all the lateral processes reduced to short curved spines, the second dorsal process 

 to a broad triangle ; another lacks seta b at one side or both ; a third lacks seta b and 



