52 Canadian Record of Science. 
those which have consolidated spicules or firm external cor- 
tex, are those most likely to retain their original forms. 
Even the looser kinds of sponges, however, may under cer- 
tain circumstances preserve their rotundity of form, in 
which case they will usually show external markings, but 
not so well internal structure, unless when sliced. On the 
other hand, when completely flattened, which is usually the 
case in shaly beds, only an outline of the form remains, and 
sometimes not even this, while the forms and in part the 
arrangement of the spicules are usually apparent. Farther, 
the hollow and thin-walled species are more liable to be 
completely flattened, though in some cases, as in the Devo- 
nian Dictyospongie, they may retain their form. It was 
this property, and the membranous appearance of the outer 
coat, that for a long time sustained the belief that these 
were plants rather than sponges. 
In the case of the sponges procured in the shales at 
Little Metis, perfect flattening has occurred, and in many 
cases the spicules have been separated, and appear as mere 
spicular patches or layers. In other instances, however, they 
remain approximately in their natural position, and even the 
general outline of the form can be observed. The collec- 
tions include several species of sponges, Hexactinellid and 
Monactinellid; but, so far as observed, one of them is more 
abundant and better preserved than the others. The fol- 
lowing may serve as a preliminary rough description of the 
species collected,—which will be more fully described and 
commented on by Dr. J. George Hinde, F.G.8., the author 
of the British Museum Catalogue of Fossil Sponges. See 
paper appended. 
1. Protospongia tetranema. S. N. (Fig. 1)' The general 
form has been spheroidal, probably with an osculum or 
oscula at top. Root composed of four long spicules in two 
pairs, which diverge somewhat and then bend toward each 
' This figure is a restoration, with two of the spicules enlarged. 
The defensive spicuies and osculum are conjectural, being based 
merely on loose spicules and general form. 
