MONOGRAPH OF THE EXISTING CKINOIDS. 81 



have nothing in common with the undivided arms in the zygometrid genus 

 Eudiocrinus^ in which IBr series are present as usual, although the second seg- 

 ment is not axillary. 



In the Macrophreata there are two instances of a curious and interesting 

 condition which occur in two widely different groups. Promachocrirms, belong- 

 ing to the Heliometrinre, and Thaumatocrinus (part 1, figs. 113, 114, p. 181), 

 belonging to the Pentametrocrinidse, have each 10 radials and consequently 10 

 post-radial series, resulting in 20 arms in the former and 10 in the latter. 



It would seem that this 10-rayed condition is unstable, for several specimens 

 of a species of Thaumatocnnus which I have examined have one or more of the 

 interradial arms absent, a large interradial occupying the position of the normal 

 interradial radial, and of the specimens of P romachocnnus kerguelensis brought 

 home by the Gauss no less than 56 per cent had less than 10 rays, most of them 

 having 6 only and one ha^ang 5 ; but no mature specimens of this species have 

 ever been found with less than 10 rays, and the normal pentacrinoids appear to 

 acquire all five additional rays practically simultaneously. 



The pentacrinoids of Promachocnnus l-erffiielensis (figs. 881-937, pp. 533-549) 

 resemble those of related species until a considerable size is reached, when five 

 interradials alternating with the five primary radials make their appearance. 

 These rapidly increase in size and give rise to postradial series in exactly the same 

 way as the original radials; but it is not until almost the full size has been reached 

 that both radial and interradial radials and arms become of the same size and 

 indistinguishable. 



Apparently the increase in the number of radials and arms in Thaumato- 

 crinus takes place exactly as in Promachocnnus ; and the process described and 

 figured by P. H. Carpenter on the posterior interradial of the young of Thaumato- 

 crinus renovatus is undoubtedly merely the rudiment of the posterior interradial 

 arm. 



The so-called interradial radials arising between the primary radials are 

 therefore situated exactly over the basals and are truly interradial in position. 



The interradial radials are apparently connected with the radials to their 

 right as viewed from the center of the disk; their skeleton is entirely independent 

 of that of the radial radials on either side, but the ambulacra! grooves and asso- 

 ciated ventral structures are derived as offshoots from those running to the radial 

 radials to the right. In the case of 6-rayed indi\aduals, therefore, the additional 

 ray is inserted behind the left posterior normal ray, just as in the case of the 

 ()-rayed specimens of Trojtiovietra picta to be later described (p. 82). 



Although in Promcwhocrinus and in Thaumatocrinus there are 10 radials 

 and postradial series, the basals and infrabasals and the rays of the basal star 

 remain 5 in number. 



I once dredged a specimen of Heliomctra. tiicuxima with 12 arms (fig. 164, p. 86, 

 and pi. 54, fig. 1348), two postradial series arising from a partially divided radial, 

 and P. H. Carpenter has recorded a specimen of Neocomatella pulchella in which 

 there were six postradial series, two of them originating on an enlarged and axillary 

 radial. These may represent a step toward the conditions found in Proma- 



