MINUTE STRUCTURE OF STROMATOPORA AND ITS ALLIES. 231 



In the third place, there remain for us to consider the possible 

 relations between the Stromatoporoids and the Calcispongige. In 

 this connexion we may first note the remarkable resemblances 

 between the Sponges and the Stromatoporoids as regards their 

 general form, their ordinary modes of growth (as evinced by some 

 forms, though not all), and the principal characters presented to 

 the unassisted eye by the free surfaces of each. We may next 

 note the resemblance between these two groups, constituted by 

 the presence in many Stromatoporoids (and perhaps in all) of 

 two sets of apertures in the superficial layer, and two sets of 

 canals perforating the skeleton. When we add that the apertures 

 of the larger canals in some forms are placed at the summit of 

 conical eminences, and that they may be disposed with some 

 regularity, we have said enough to show how striking is the 

 general resemblance between these two sets of openings in the 

 Stromatoporoids and the "pores" and " oscula" of the Sponges. 

 On the other hand, to resemblances like the above, which have 

 so strongly impressed many observers, we have to oppose some 

 serious and important points of dissimilarity. By far the most 

 important of these is to be found in the nature of the skeleton. 

 In all the known Calcispongise, the skeleton consists of free 

 spicules, which are never amalgamated or fused with one another, 

 and which, therefore, never form a continuous framework, how- 

 ever densely they may be packed together. In no Stromato- 

 poroid, however, have free calcareous spicules ever been detected 

 by any observer ; and the skeleton is undoubtedly a more or less 

 continuous one. If, therefore, we were to accept the view that 

 the Stromatoporoidea were really referable to the Calcareous 

 Sponges, we should have to assume that they constitute a special,, 

 peculiar group, bearing a relation to the typical Calcispongise 

 somewhat similar to that which the Lithistidge or Hexactinellidse 

 of the present day bear to those Siliceous Sponges in which the 

 spicules are not united or soldered together (Sarcohexactinellida 

 of Carter, the Lyssakina of Zittel). Nor can any serious a priori 

 objection be brought against this view, since there is no funda- 

 mental reason why there should not be a group of Calcispongise 

 with free spicules, and another group of the same with a vermicu- 

 late or reticulate continuous skeleton, just as we know that in 

 nature there are found parallel groups amongst the Siliceous 

 Sponges. 



But withal there is still the further trenchant difficulty to be 



