REPORT ON THE ECHINOIDEA. 15 



Structural importance assigned to them by Mr Mackintosh, and when we come to include 

 within his classification the spines of the Spatangoids and Clypeastroids, according to the 

 structural features he has employed to separate his principal series, we should be com- 

 pelled to unite into one series groups which have no systematic affinity and are zoologi- 

 cally widely separated. The mere fact that the spines of the Diadematidje are hollow 

 does not seem a sufficient reason for contrasting them to the spines of all the other 

 Desmosticha. I should be more inclined to consider the spines of the Diadematidse 

 (adopting the nomenclature of Mackintosh) as monocyclic Acanthosphenota, with a more 

 or less hollow interior. An excellent example of the type showing affinities to the 

 Diadematid^ and to the Echinidse is Pseudoholetia ; in one section (PL XXXVIII. fig. 3) 

 we have the hollow spine as in Diadematidae, in the other (PI. XXXIX. fig. 11) the 

 central portion of the shaft is completely occupied by reticulations as in the Echinidae. 



From the examination of the few young spines of Echinids which have been figured thus 

 far either by MtiUer^ or by myself,^ there does not seem to be in the early stages very great 

 differences in the structure of the spines. The young spines are in all cases polygonal, 

 made up of rectangular meshes placed in regular stories one above the other ; the upper 

 set of meshes open, while the outer beams send off into the interior smaller rods, the first 

 rudiments of the second or third row of wedges of the polycyclic spines or merely lateral 

 offshoots connecting the large calcareous wedges, the original beams forming the 

 rectangular meshes of the young Sea-urchin. There is no difference in the typical 

 structure of the spine of the young of Cidaris, Echinus, Strong ylocentrotus, Arhacia, 

 Echinocyamus, or Schizaster, the genera of which the young spines have thus far been 

 figured. 



The modifications which eventually give to the spines their final characteristics are 

 all derived from the changes undergone by this single primitive fenestrate type, and 

 are not features which are found developed eai-ly in the plutean stage, or based upon radi- 

 cally different types of structure. The very fact that we have among the Echinids the 

 anomalies to which Mackintosh refers, shows us plainly that the derivation from the 

 original embryonic type has not gone on during growth equally in all the genera of the 

 same family, some of the genera retaining a much more embryonic condition than others. 

 By embryonic, I mean the simple fenestrate structure of the spines such as still exists 

 among some of the Clypeastroids and Sj^atangoids, in which the reticulation docs not 

 extend towards the central part of the shafts, or in the simpler monocyclic types of such 

 genera as Astropyga and the Salmacidse. It is among the Desmosticha, belonging to the 

 Cidaridse, the Salenidse, the Diadematidse, and the Echinidaj, among the oldest genera, 



1 J. Miiller, Ueber die Larven n. Metamorphosen der Opliiuren u. Seeigel, Abband, Berlin Akad., 1848-1855, 

 Abbandlungen, i., iv., yi., vii. 



- A. Agassiz, Embryology of Echinodenns, Mem. An. Acad., 1864 ; The Hoimdogies of Pedicellai:£e, Am. 

 Nativralist, 1873. 



