PROPAGATION OF CEKTAIN ECHINODERMS. Gl 



stages are run through rapidly, and that the young are passed 

 back from the ovai'ial opening, which is at tlie side of the mouth, 

 along the dorsal ambulacra, and arranged in their places by the 

 automatic action of the ambulacral tentacles themselves. 



This is one of the cases in which there is no special marsupium 

 formed ; it is possible that the comparatively genial conditions of 

 the land-locked fiords and harbours of the Malvinas, and the ad- 

 ditional shelter yielded by the imbricating fronds of Macrocystis, 

 may render such exceptional provision unnecessary. 



On the morning of the 7th of February, 1875, we dredged at a 

 depth of 75 fathoms, at the entrance of Corinthian Harbour {alias 

 " Whiskey Bay ") in Heard Island (so far as I am aware the most 

 desolate spot on God's earth), a number of specimens of a pretty 

 little Psoitis, which I shall here call for the sake of convenience 

 P. epMppifer, although it may very possibly turn out to be a 

 variety of the northern P. operculatus. 



P. epMppifer (figs. 2, 3) is a small species, about 40 millims. in 

 length by 15 to 18 millims. in extreme width. In accordance with 

 the characters of the genus, the ambulatory area is abruptly de- 

 fined, and tentacular feet are absent on the upper surface of the 

 body, which is covered with a thick leathery membrane in which 

 calcareous scales of irregular form are imbedded. The oral and 

 excretory openings are on the upper surface, a little behind the 

 anterior border of the ambulatory tract, and a little in advance of 

 the posterior extremity of the body, respectively. A slightly ele- 

 vated pyramid of five very accurately fitting calcareous valves 

 closes over the oral aperture and the ring of oral tentacles ; and a 

 less regular valvular arrangement covers the vent. 



In the middle of the back in the female there is a well-defined 

 saddle-like elevation formed of large tessellated plates somewhat 

 irregular in form, with the surfaces smoothly granulated (fig. 2). 

 On removing one or two of the central plates we find that they are 

 not, like the other plates of the perisom, imbedded partially or 

 almost completely in the skin, but that they are raised up on a cen- 

 tral column like a mushroom or a card-table, expanding above to 

 the form of the exposed portion of the plate, contracting to a stem 

 or neck, and then expanding again into an irregular foot, which 

 is imbedded in the soft tissue of the perisom ; the consequence of 

 this arrnagement is that when the plates are fitted together edge 

 to edge, cloister-like spaces are left between their supporting- 

 columns. In these spaces the eggs are hatched, and theeggs or the 



