ECHINOIDBA. II. loi 



strongly serrate at the edge (PI. XV. Fig. 35). Tlie sphæridiæ do not present prominent features; they 

 occur (in the larger specimens) also at the large tubefeet at the posterior end of the test. 



It is an iniportant faet tliat even in the smallest specimens there is no trace of a latero-anal 

 fasciole, so that it may be regarded as proved that this fasciole is never fonnd in Hciiiiastcr — a very 

 characteristic difference from the yonng of the geuns Abatus. In the yonng Abatns there is a large 

 fasciole enclosing botli the apical system and the anal area; a transverse band then develops between 

 the apical and anal area, and the part of the original fasciole behind the transverse band thus 

 becomes the latero-anal fasciole, whereas the anterior part of the original fasciole in connection with 

 the transverse band forms the peripetalons fasciole. In llciniastcr the anal area is never intrafasciolar.' 

 In the specinien of 3'""' the peripetalons fasciole is already distinct (Fig. 17), and at a comparatively 

 large distance from the anal area. It is very small and in the anterior petal only one tnbe-foot is 

 distinct — and by no means very large — and two more are about to appear. In specimens a little larger 

 the peripetalons fasciole is very prominent, broad, bnt still enclosing only a very small space (PI. IV. 

 Fig. 10); upon the whole the fasciole is comparatively much broader in the smaller specimens. The 

 odd anterior ambnlacrnm develops early, thus at a size of 5 — 6""" already 4—5 rather large tube-feet 

 are formed. The paired petals are not developed till later on. In a specimen of 10""" length I find in 

 the antero-lateral petals 5 pairs of pores in each series, but of the postero-lateral petals no trace is 

 seen as yet. In a specimen 12""" in length I find 2 pairs of pores in eacli series in the j^ostero-lateral 

 petals. The smallest specinien. in which I have found the genital pores developed was 14""" long. 



This species was taken by the v; Ingolfs at the following stations: 



St. 24 (63° 06' Lat. N. 56° 00' Long. W. 1199 fathonis 2'4C. Bottom tenip.) i specimen. 



Unfortunately several of the specimens were in a more or less broken condition. — The spe- 

 cies was further taken by the Thor at 62° 57' Lat. N. 19° 58' Long. W. 975 M. (1903) and by «Michael 

 Sars>, 61° 40' Lat. N. 3° 11' Long. IC. 220 fathom.s, 6°3 bottoni temperature (Ad. Jensen. 1902). The latter 

 locality (the Shetland-Norway ridge) is rather surprising and may indicate the possibility of the sjjecies 

 occurring along Norway. (Comp. Ecliiiius Alcxandri). 



The geographical distribution of this species is thus the Northern Atlantic, from the Davis 

 Strait to the Caribbean Sea and from South of Iceland to the Azores. It belongs to the w-arni 



I This feature, combiued with the ethmophract apical system, the 4 genital pores, the tliffereiice in the pedicellariæ 

 (evidently the least important character) and the very great difference in the whole shape and appearance, proves beyond 

 douVjt that Loven was quite right in maintaining that the antarctic forms: Abalus cavcriiosus etc. cannot be referred to the 

 genus Hcmiaster, as done by Professor .\gassi7,. «An extraneous form Hke this, if suffered to remain in the otherwise homo- 

 geneous group of true Hemiasters, is sure to vitiate its integrity, and the nuxed a.ssemblage thus set up for a natural genus, 

 if taken on trust, cannot fail to mislead when the question is to trace out coniparativelj- its former geological and actual 

 geographical distributions. iLovén. On Povirtalesia. p. 73). In his last work, «The Panamic Deep-Sea Echini >, Agassiz 

 recognizes the correctness of Lov én 's views, while Doderlein (Echinoidea d. deutsch. Tiefsee-Exp.) still refers the antarctic 

 forms to Hemiaster, without entering upon the question, however. This question will be treated at more length in my Re- 

 ports on the Echinoidea of the German and Swedish South-Polar Expeditions). 



