126 NATURAL SCIENCE Pee 
deposit of calcareous tissue takes place between the basal ectoderm 
and the surface of attachment, and that it is secreted by, not formed 
within the ectoderm cells. These statements are confirmed by H. V. 
Wilson for Manicina areolata, and in a question of this sort it requires 
very strong evidence to upset the proofs from embryology. Miss 
Ogilvie’s evidence is hardly strong enough; the appearances which 
she describes are not unfamiliar to students of corals and are suscep- 
tible of a different interpretation, but she has at least reopened the 
question, which will have to be settled on better evidence than that 
which she has adduced. 
Amongst the many figures which illustrate the work there are 
several showing the relations of hard and soft parts in recent corals. 
Some of them are correct, others are misleading, if not positively 
incorrect. Take, for instance, the diagram of Zurbinaria on p. 209. 
The anatomy of this genus has been thoroughly described by Dr 
Fowler, and we learn from him that there is a system of canals which 
permeate the corallum and communicate with the polyp cavities. 
These canals anastomose freely, but Miss Ogilvie’s figure shows only 
a few digitate or branched diverticula; no anastomoses, no network, 
and no transverse communications with the polyp cavities. The 
figure of Fungia on p. 169 can only be called a diagram of theoretical 
relations. Asa matter of fact the soft parts of Fungia have not the 
structure shown in the figure. Miss Ogilvie homologises the tissues 
on the aboral face of Fungia with the edge-zone of other corals. This 
is right enough, but it is not right to assume, as she does, that there 
is no communication between the synapticula and through the theca 
between the cavity of the edge-zone and the general cavity of the 
body. Asa matter of fact definite canals pass between the synapticula, 
some are united below the level of the synapticula by a radial canal, 
some are directly continuous with the cavities of the edge-zone. The 
mesenteries are best developed above the synapticula, but some extend 
also between and even below them, the rule being that the mesenteries 
are attached to synapticula, either above or at their sides, but some 
extend far down and send narrow mesogloeal bands to be attached to 
the basal wall of the disc. Other features, such as the position of 
the tentacles, are not correctly represented. The writer has the more 
confidence in making these statements since he has recently examined 
the anatomy of Fwngia, in order to test the correctness of this figure. 
One is inclined to suspect that Miss Ogilvie, whilst making plentiful 
-use of the anatomical researches of other authors, has not herself any 
ereat familiarity with the structure of coral polyps. There is some 
internal evidence that, after deciding in her own mind how the 
corallum was formed, she has inferred the anatomy of the polyps from 
the microscopical characters of their coralia, without studying the 
actual relations in a sufficient number of instances. Such inferences 
are apt to be misleading. Whether this is the case or not, Miss 
Ogilvie has been led by her views on the formation of calcareous 
tissue to give a lively but an unlikely account of madreporarian 
development and evolution, an account which is in harmony with the 
figures criticised above, but which does not and cannot explain the 
diagnostic character of the Madreporaria Perforata of M. Edwards and 
Haime—viz. the presence of a complex canalicular system in the wall, 
