176 THE VOYAGE OF H.M.S. CHALLENGER. 



deep pit iu which the anal system of Lovenia elongata is placed. The actinal plastron is 

 quite small, and the primary tubercles with sunken serobicular areas forming the adjoining 

 pavement, and the outer edge of the bare actinal ambulacral areas are comparatively 

 smaller than in Lovenia elongata and more closely packed together (PI. XXXV.'' fig. 7). 



The large primary tubercles which in Lovenia elongata extend from the ambitus to 

 the abactinal res-ion in the anterior interambulacrum, and in the anterior half of the 

 lateral posterior ambulacra, are in Lovenia suhcarinata limited to two horizontal rows of 

 two and three large primaries (PI. XXXV.'^ fig. 5) which alone carry stout, long curved 

 spines on the abactinal surface. The rest of the abactinal side of the test is covered by 

 thin, short, slender curved spines, while in Lovenia elongata the large curved spines 

 extend over the greater part of the posterior extremity of the test (see pi. xix.'' Eevis. 

 Ech.). The tufts of spines on the two sides of the plastron enclosed by the subanal 

 fasciole consist of somewhat longer spines than those of the abactinal surface, and these 

 tufts consist of much more slender spines than those forming the corresponding tufts of 

 Lovenia elongata. The primary spines of the actinal surface are also shorter and com- 

 paratively more slender than in Lovenia elongata. 



This species is also interesting on account of its rudimentary lateral fasciole, which 

 extends close to the ambitus from the anterior ambulacrum to about the median line of 

 the posterior lateral ambulacra. This fasciole is somewhat indistinct, but consists of two 

 to three more or less irregular horizontal lines of small miliary tubercles, differing in no 

 way from the smaller miliaries covering other parts of the test (see Duncan). This seems 

 to show quite conclusively a far closer relationship between Breynia and Lovenia than 

 had been suspected. In fact, if we are to take Lovenia elongata as the t}^Dical Lovenia 

 and Breynia australasice as the typical Breynia, Lovenia suhcarinata has, like Lovenia 

 elongata, the large primaries with sunken serobicular areas of the actinal surface, and the 

 same arrangement of the primary tubercles of the anterior part of the abactinal part of 

 the test, while it has the anal system and the whole of the posterior part of the test more 

 like Breynia australasice, and in addition the rudimentary lateral fasciole, the remnant of 

 the peripetalous fasciole of Breynia, which has been considered one of the principal points 

 of difference between it and Lovenia. 



The existence of a partial lateral fasciole, both in Lovenia elongata and iu Lovenia 

 cordiformis, if we may so call the somewhat irregular band of miliaries extending from 

 the anterior ambulacrum on the abactinal side near the edge of the ambitus towards the 

 posterior extremity, throws considerable light on the origin of the fascioles, and plainly 

 shows that they are at first (at least iu Lovenia and Maretia where lateral fascioles had 

 not been observed) more or less irregular bands of miliary spines, which eventually 

 become specialised and limited to distinct areas. The origin and formation of the sub- 

 anal fascioles, as well as that of the peripetalous fasciole wherever I have traced it, fully 

 sustains this view. This helps to explain the great variation we find in the degree and 



