14 



spines which present characters intermediate in some ways between the other two". 

 This is quite true, but it may be remarked that the statement holds good for all the 

 species o{ Diademet. In all of them the primary spines of the actinal side are somewhat 

 fusiform or clubshaped, indistinctly verticillate, especially at the point, longitudinally 

 striated; while towards the abactinal side they gradually assume the typical form. 



The living animal atTords a most beautiful sight, with its well known blue 

 spots, contrasting with the black colour of the test. The anal tube is about 1 cm. long, 

 grey, with a band of yellowish red around the opening. The colouring of the 

 spines (black banded or even quite white^ is of no specific importance, as is also 

 asserted by most previous authors. The small secondary spines are in a constant, 

 lively circular motion, the large spines being much more slowly moved. — The 

 habits of this most interesting sea-urchin are well known (Sahasin, Dödf.ulein, 

 Sav. Kent etc.), and my observations agree with those of these authors. The symbiosis 

 of fishes (Engraulis?) ^Nith this species (Sarasin "), Coutière''^)) I have not observed. 



Numerous specimens were found at Koh Chang and Koh Kahdat; mostly 

 they were seen in large numbers (c. 10 — 20) together in bare places between the 

 corals, at a depth of 1 — 5 fathoms. Some specimens were taken at Pulo Redang 

 by the Skeat Expedition. 



The distribution of D. saxatile has hitherto (since Rev. ofEch.) always been 

 said to be cosmopolitan in the warm region. As now the Atlantic form proves to 

 be a distinct species, the occurence of the true D. saxatile is restricted to the Indo- 

 pacific ocean. Here it seems to be universally distributed in the warm region 

 (together with D. Savignyi). I have seen specimens from Zanzibar, Singapore, Siam, 

 Cebu and Tahiti. From the West-American coast it is not known. 



I shall now point out the characters by which the other species of Diadema 

 are distinguished from D. saxatile. 



D. autillarnm Phil. The pore areas are not distinctly widened at the 

 actinostome and the pores here not much smaller than those above the ambitus. 

 As in D. saxatile there is not a primary tubercle on each compound ambulacral 

 plate; in small specimens, however, there is a primary tubercle on each plate 

 and such is evidently the case also in D. saxatile. It seems then that some of 

 these tubercles — and spines — are absorbed as growth proceeds. The uppermost one of 

 the large interambulacral tubercles of the inner series is found on the 4th or 5th plate 

 from above, the naked space being smaller and the inner series of tubercles reaching 

 higher above the ambitus than in saxatile. There are no dark impressions on the geni- 

 tal plates. Apical system small (see measurements given below). The triphyllous 



') Über einen Lederigel aus dem Hafen von Trincomalie (Ceylon) und seinem Giftapparat. Zool. 

 Anzeiger IX. 1886. p. 82. 



') Observations sur quelques animaux des récifs madréporiques de Djibouti. Bull. Mus. d'hist. nat. 

 Paris. 1898. p. 238. ' 



