Prof. P. M. Duncan on the Salenidse. 251 



Other long spines, shorter than the above, are more slender, and 

 the whorls of spinules are not so numerous ; but there is still a 

 faint pigmentation on and just above the milled ring. A 

 third set of spines are shorter and more slender than these 

 last ; and a fourth, noticed close to the peristome, in the ambu- 

 lacra and just above the sphferidia, are still smaller, slightly 

 bent, with or without whorls of spinules, and with the pig- 

 mentation carried along further on the longitudinal ridges 

 (fig. 6). 



The next group of spines are very remarkable, and thej are 

 all short. Some are large at the milled ring, and stout and 

 conical, looking like aborted great spines; they are marked 

 with lines of pigment to the tip. Others are smaller and 

 have serrate edges to the ridges and the same kind of orna- 

 mentation (fig. 5). In all these kinds of spines the milled ring 

 is large and faintly crenulate. 



Finally there are numerous flat club- or stout spatula- shaped 

 short spines on the ambulacra and median interambulacra. 

 These are situated on small tubercles, which are not crenulate ; 

 and the milled ring of each spine is a mere expansion : 

 usually the shaft is constricted for a space and then expands 

 gradually and rounds off close to the tip (fig. 4). Some are rod- 

 shaped (but this is not common) ; and some are bent laterally ; 

 and in all, lines of pigment run longitudinally and conform to 

 the outline, having a dotted appearance here and there. Some 

 of these spines are found on the anal membrane, which covers 

 the plates there (fig. 2) ; and they are evidently connected with 

 minute tubercles. A vertical row of these spines, many 

 being small and thin, covers the tubercles just outside the 

 poriferous zone in the interambulacra. Finally, above the 

 ambitus it will be noticed that in one or two places the club- 

 sha])ed spines do assimilate in position to those of the Cidarida?, 

 and environ more or less the bases of the great tubercles. The 

 vertical rows of these spines, both in the ambulacra and in the 

 median space of the interambulacra, diverge at their tips, and 

 present the appearance so Avell illustrated by A. Agassiz 

 (Hg.4). 



The comparison of these spines and those described by 

 A. Agassiz with the fossil forms will be made in the next com- 

 munication ; but it is as well to remark that the reseml)lance 

 is only in the variety of the sj)ines in a slightly clubbed con- 

 dition in the fossils and in some of them being bent. 



The abactinal system of Salenia inofundi differs materially 

 from the form described by A. Agassiz, and less so from 

 Hdhjiia Goes'iana. I agree with Lovcn, Wright, and Cotteau, 

 and have always taught, that the position of the madreporic 



