1900.] ME, F. P. BEDFORD OX MAI/A.YA3T ECHINODERMS. 275 



coral, sand-grains, &c, but I have obtained it on the " tangles " 

 from mud. 



Distribution. The area of distribution is a somewhat anomalous 

 one : it is fairly widely distributed over the Indo-Pacific circum- 

 tropical zone, extending from the Coromandel Coast on the west 

 through the Bay of Bengal and Malay Archipelago to the Prince 

 of Wales Is. : in the east and southwards it extends into the 

 Antarctic littoral zone, having been found in Tasmania as well as 

 on the South-Australian coast, but it is not reported from as far 

 north as Hongkong or the S. Japanese seas. 



There seems little doubt that M. de Loriol is right in associ- 

 ating Studer's Schleinitzia crenularis from New Guinea with this 

 species. Prof. Troschel and others have shown the inconstancy 

 which the tubercles exhibit with regard to absence or presence 

 of crenulation among recent Cidaridae ; and I have followed 

 M. de Loriol (loc. cit. 1883, p. 1) in uniting Agassiz's species of 

 Phyllacanihus and Stephanocidaris bisjyinosa iu Desor's genus 

 Rhabdocidaris. 



The specimens brought home represent a series showing growth- 

 changes from an individual 8 mm. in diameter to one 57 mm. in 

 diameter. Considerable allowance has to be made for individual 

 variations in drawing deductions from so few specimens ; but it 

 seems worth while to give a table of the principal measurements 

 and also to figure the changes taking place in the spines during 

 growth. 



As is well known, new coronal plates are formed between the 

 calycinals and the corona, the first-formed appearing as minute 

 sectors between the oculars and genitals (c/. condition in 

 Asthenosoma) ; and Prof. Loven has described the process of 

 formation of the perignathic girdle and peristomal radial and 

 interradial plates by absorption from the coronal plates and 

 redeposition on the buccal membrane. In this species, unlike 

 Dorocidaris papillata examined by him, the peristomal plates do 

 not increase in number after the animal has reached a diameter of 

 22 mm.; and judging from the relative positions of the plates 

 and spines of the corona bordering on the peristome, I conclude 

 that no entire coronal plates have been absorbed in that region 

 subsequent to the stage represented by the smallest of our 

 specimens (8 mm. in diameter). The formation of new coronal 

 plates probably continues throughout life, and the spines 

 evidently increase in length up to a quite late stage ; but whether 

 they keep pace with the growth of the test is not certain, as in the 

 largest specimen they had apparently been injured during life. 

 When first formed, the spines are always smooth and acquire the 

 characteristic granulations and thorns later ; but it is interesting to 

 note that in the smaller specimens the spines acquire the typical 

 characteristics long before they reach a length at which granu- 

 lation begins to appear in the spines of larger individuals (cf. 

 Plate XXI. figs. 1 b, 1 d). In some specimens all the tubercles are 

 smooth, while in others many of the larger tubercles are very 



