128 DR A. MILNES MARSHALL ON THE 



iu the adult form the smallest amount of difference between the constituent 

 individuals ; while those forms in which this differentiation reaches its greatest 

 development will be the most highly modified forms. These principles are of 

 great importance in framing a scheme of classification of a colonial group such 

 as the Pennatulida, and have not received sufficient attention in the classifica- 

 tion at present in use. 



In the ordinary P. phosphorea the amount of differentiation is comparatively 

 slight, and is brought about in the simplest possible manner ; the asexual zooids 

 being simply arrested at what is merely an early stage of development in the 

 case of the polyps. This is well shown in PL XXI. fig. 7, representing a 

 side view of the lower end of the rachis, and showing the early stages of 

 development of the polyps and zooids. 



The figure shows that the young polyps d are at first quite independent of 

 one another, and that in their earlier stages they are absolutely identical with 

 the zooids e ; and that the differences arise from the zooids becoming arrested 

 at this early stage, while the polyps advance further, increase in size, acquire 

 calyx processes /, and tentacles t, fuse with one another at their bases, so that 

 their further increase in length gives rise to the leaf of the adult, and acquire 

 the full number of mesenterial filaments, and ultimately reproductive organs. 



The differentiation is thus of the simplest possible character, the zooids 

 being simply arrested or abortive polyps, whose function is apparently to main- 

 tain currents of water circulating throughout the colony, for which purpose they 

 have retained the sole structure peculiar to them — the clothing of exceptionally 

 strong cilia on the abaxial surface of the stomodseum. As the siphonogiyphe is 

 present in the mature polyps of many Alcyonarians, such as Alcyonium (vide 

 Hickson, he. cit.), it seems certain that in the Pennatulida it is a structure that 

 has been lost by the polyps, but retained by the zooids. 



In the variety aculeata differentiation has advanced further; and it is a 

 point of importance to note that the points in which the large zooids differ 

 from the small ones cannot be considered as repetitions of any part of the 

 process by which the polyp is developed from the zooid condition. In the 

 young polyp all the calyx processes arise simultaneously (fig. 7), as do also all 

 the tentacles, so that the asymmetrical development of the calyx in the large 

 zooids must be regarded as peculiar to and acquired by them. The lateral 

 position of the mouth in the large zooids has apparently been acquired inde- 

 pendently of and previous to the formation of the calycular spine, inasmuch as 

 it is an equally prominent feature in the normal small zooids (fig. 8, e). 



Judging from their structure, the large zooids would seem to be protective 

 in function, but as to the special circumstances which determine their develop- 

 ment in particular forms, we are in complete ignorance. Were our knowledge 

 confined, so far as these forms are concerned, to the " Triton " specimens, we 



