THE BKANCillNG-SYSTEMS OF THE ALCYONACEA. 535 



surface of the mesoglcea, are not occupied by spicules, but by groups of rounded or 

 irregular colls which, as may be seen at scb. in figs. 22 and 24, are clearly derived from 

 proliferating ectoderm-cells. In sections tliat have been double-stained with saffranin or 

 picro-carmine and picro-nigrosin the mesogloea is coloured blue, which takes a deeper 

 or lighter tint according as the structureless matrix is of firmer or softer consistency. 

 Pigs. 22 and 2i< show that the mesogloea near the surface is full of spaces and is in most 

 ■parts soft and lightly stained. Eurther, that not only are the spaces generally occupied 

 by cells derived from the ectodenu, but that processes of mesoglcea, marked tiir/', appear 

 to stretch outward between the ectoderm-cells themselves and to isolate individual cells 

 or groups of cells. In fact, wo have here the mesogloea in course of formation as an 

 intercellular substance. The ectoderm-cells proliferate, and while proliferating they 

 secrete the jelly-substance of the mesogloea. Eventually a honeycomb of mesogloea is 

 formed, in the cavities of which groups of proliferated cells are enclosed. These enclosed 

 cells are the scleroblasts, and as they become more deeply situated they give rise to 

 spicules in the usual manner. In many cases I have found minute spicule-sheaths, 

 stained deeply by saflfranin and with traces of the organic axis of filaments in their 

 interiors, among the proliferating cells just l)olow the surface of the ectoderm, but they 

 are so minute that I have not been able to figure them satisfactorily. It is clear, then, 

 that the mesogloea increases in thickness during the growth of the anthodete as a result 

 of the activity of the ectoderm-cells, and that, as it increases, scleroblasts, budded off 

 from the ectoderm, are enclosed in it and give rise to spicules. I have described a 

 similar origin of the mesogloea and scleroblasts in Heliopora cccrulea *, but in this case, 

 as no spicules are formed, the subsequent history of the scleroblasts is different. 



The endoderm in the deeper parts of my specimens was so much macerated that I can 

 say nothing definite about it ; but it was generally fairly well preserved in places 

 where new growth was taking place, especially in young buds. In all the superficial 

 parts of the anthodete zooxanthelhe occur al:)undantly in the endoderm, aiid wliere the 

 latter has undergone extensive maceration they lie loose in the coelenterou. The best 

 preserved parts of my specimens indicate that the endoderm is much vacuolated, and 

 that the zooxanthelli^ are contained in tlie endoderm-cells. In some cases (see fig. 23, 

 zx^ loose zooxanthellae are contained in cells which may be Avandering cells. Zooxan- 

 thellfe are specially abundant in the cavities of the tentacles aiul their pinnules, and the 

 more one studies Alcyonarians the more one is strixek by the accumulation of these 

 organisms in the parts most exposed to the light. One very rarely finds any traces of 

 food in the coelentera of Alcyonarians ; and its absence and the accumulation of symbiotic 

 zooxanthelkc in the nutritive layer, the endoderm, point to the conclusion that the 

 Alcyonaria are, in many cases at least, truly symbiotic organisms depending chiefly, if 

 not exclusively, on the zooxauthelkie for their nutrition. Mr. Stanley Gardiner's f 

 recent and conclusive experiments on the evolution of oxygen by corals living expanded 

 in sunliglit gives weighty support to this view. 



* Philosophical Transactions, vol. clsxxvi. (ISiJo). 



t Proo. Camhridge Philosophical Society, vol. ix. (189-<», pp. -irS2-4S4. Mr. (jardiiar, however, Joes not attribute 

 ;the evolution of oxyi;en wliolly to the zooxaiithelhc. 



