REPORT ON THE ECHINOIDEA. 125 



mesenteric fold somewhat similar to that of Rhynchopygus, but not extending like a 

 diaphragm across the test. In Spatagocystis, the course of the alimentary canal (PI. 

 XXVL* fig. 1) is much the same. It is however more free within the test, and not 

 supported by so many mesenteries. The intestine also opens in the posterior part of the 

 pouch (PI. XXVI. ^ fig. 5), forming the anal system (PI. XXVI." fig. 6). In Pourtalesia 

 proper, although the anal system is deeply sunken, there is no such pouch formed, the 

 plates of the anal system are nearly on a level with the surface of that part of the test 

 (PI. XXII. fig. 14; PI. XXII." fig. 13; PI. XXVIII." fig. 6; PI. XXXV. » fig. 13). In most 

 of the species of Pourtalesia proper, the course of the alimentary canal is still less definite, 

 althouoh havinof in -oreneral the trend of the genera noticed above, as is well shown in the 

 profile views of two specimens of Po?«taZesia hispida (PL XXII. figs. 10, 11), which have 

 the rounded alimentary canal of Spatagocystis^ and no prominent mesenteries. The 

 intestine is well defined in this species (PL XXII. fig. 14) as well as in Spatagocystis 

 (PL XXVL* fig. 6). Hoffman's organ is present in the Pourtalesise, and is well shown 

 (PL XXII." fig. 12) in Pourtalesia laguncula. In Pourtalesia ceratopyga we have again 

 the broad alimentary canal well supported on the test by mesenteries, much broader 

 even than we find them in Echinocrepis. 



The actinostome of the Pourtalesise is elliptical, forming the large opening of the 

 actinal groove. It is covered by a membrane strengthened by an outer row of plates 

 (PL XXVIII." fig. 9), from the centre of which leads the narrow cesophagus (PL XXIL* 

 figs. 12, 15) ; it is not labiate as in Schizaster and allied genera, but the structure 

 of the actinostome is more closely allied to that of the Spatangoid genera having the 

 edges of the actinostome in one plane. The structure of the actinal groove is due, as is 

 well seen in a profile view, to the enormous development of the odd anterior ambulacrum, 

 the plates immediately adjoining the actinal edge being fully as large or larger than 

 those of the adjoining interambulacra (PL XXII. figs. 15, 17; PL XXII." fig. 15; 

 PL XXVIII." figs. 9, 11), while those of the anterior lateral ambulacrum are smaller 

 and those of the posterior lateral ambulacra immediately adjoining the actinostome are 

 extremely narrow and elongate (PL XXII." figs. 2, 9 ; PL XXVIIL-fig. 10 ; PL XXXV.- 

 fig. 10 ; PL XXVI." figs. 2, 4). This is of course in entire opposition to the usual structure 

 of the actinal region in Spatangoids, in which owing to the position of the actinostome coin- 

 cident with the general level of the actinal region and the uniformity of the small .size of 

 the plates of the ambulacral areas immediately adjoining the actinal opening as in Plate 

 XXXV. figs. 11, 12, and Plate XXXV." fig. 1, and other normal Sj)atangoids, no such difier- 

 encein the size of the ambulacral plates of the different areas exists. The first trace of this 

 gradual increase in size of the plates of the odd ambulacral area is well shown in the genus 

 Cionohrissus (PL XXIII. fig. 9), in which we have a slight actinal groove. We find the 



1 In Cystechinus the course of the narrow alimentary canal is more complicated (PI. XXIX."" fig. 6), recallin<,' 

 somewhat the course it takes in the Desmosticha. 



