104 BRITISH STROMATOPOROIDS. . 



above-described radial tubes, while other specimens have every branch full of them. 

 It is obvious, therefore, that the presence or absence of these radial tubes cannot 

 be used as a generic, or even a specific character. It is an individual peculiarity 

 with which we have to deal ; and the tubes in question can therefore only be 

 structures which are occasionally developed. The only structures, however, in 

 an ordinary Hydroid colony which are present in some individuals of a species and 

 not in others are the reproductive zooids. Thus, if we accept the conclusion that 

 the embedded radial tubes of I. oculatum belong to the organism in which they 

 occur, we are apparently shut up to the further conclusion that they must have 

 served for the lodgment of the reproductive zooids. On this view, those speci- 

 mens of I. oculatum which are destitute of these radial tubes represent the sterile 

 colonies, while the more numerous " Caunoporoid " examples are the fertile indivi- 

 duals of the species. Upon the whole, therefore, while fully admitting the diffi- 

 culty of anything like definite proof on the point, it seems to me that the most 

 probable hypothesis as to the embedded tubes of Idiostroma oculatum is to regard 

 them as connected with the function of reproduction, and as corresponding with 

 the differently constructed " ampulla? " of the recent Stylasterids. 



The only other species of Idiostroma with which I am personally acquainted is 

 the form, previously alluded to, which I have dubiously identified with the Tragos 

 capitatum of Goldfuss, and which I think is probably the Parallelopora Goldfussi 

 of Bargatzky. This type occurs commonly in the Paffrath district, and is also not 

 rare in the Devonian Limestones of Devonshire. It differs from I. Roemeri and I. 

 oculatum in not being cylindrical or fasciculate in form, but in being massive or 

 sub-massive, generally more or less spherical. Moreover, in place of a principal 

 axial canal, giving off lateral tabulate branches, we find in this species numerous 

 large tabulate tubes distributed irregularly through the ccenosteum, and quite dis- 

 tinct from the normal but also tabulate zooidal tubes. In this species, we find, 

 as I have formerly described, numerous lenticular or oval vesicles of comparatively 

 large size, scattered through the general skeletal tissue (Fig. 8), and it may be 

 conjectured that these are also of a reproductive nature, and correspond with 

 the " ampulla? " of the Stylasterids. These vesicles are generally from one 

 to three mm. in diameter, and are often crossed by internal partitions or tabula?. 

 They are often only bounded by the general skeletal tissue of the ccenosteum ; but 

 at other times they appear to have a thin proper wall of their own. Other speci- 

 mens of this species exhibit somewhat similar cavities which are surrounded by 

 greatly thickened walls ; but I have not been able to make out whether these are 

 a still further modification of the supposed " ampulla? " just spoken of, or whether 

 they are not rather embedded adventitious structures. 



