ZOOLOGICAL CONSIDERATIONS. 15 



ultimate structure as ValvuUna that is to say, for the most part subarenaceous and 

 imperforate, though often thin-walled, but also show a striking similarity in the interior 

 subdivision of the segments. 



We have,. then, in the Carboniferous fauna these four genera, embracing in their 

 modifications a very remarkable series of forms, occupying a position between the two 

 great sub-orders into which the Foraminifcra are divided, not rightly belonging to the 

 IMPERFORATA, if the definition be strictly read, though in close affinity to the most 

 strikingly imperforate types, but equally removed from the PERFORATA. 



These intermediates, whether amongst individuals, species, genera, or larger groups, 

 furnish the test under which artificial schemes of classification break down ; but if the object 

 in view be to trace the natural sequence of forms, rather than to establish a system of defini- 

 tions, the evidence they yield is precisely that which is most valuable. It has been suggested 

 that "the progress of knowledge will eventually break down all sharp demarcations, and 

 substitute series for divisions" in zoological classifications j 1 and if it be so, as indeed can 

 scarcely be doubted, intermediate forms have a significance too important to be ignored 

 merely for the sake of upholding the current definitions of existing groups. Not that this 

 need alter, at any rate in the present case, the general mode of treatment. The division 

 of the Foraminifera into " Imperforata" and " Perforata" is exceedingly convenient, and in 

 the main rests on a sound natural basis ; and if increased observation tends more and more 

 to break down this boundary-line in common with other sharp demarcations, any alternative 

 that could be proposed in the existing state of knowledge would be open to objection of 

 the same kind, varying only in degree. So long, therefore, as their true relation to the 

 series is understood, it is not very material on which side of the line at present recognised 

 the transition-group that has been described is placed ; and accepting as a guide the posi- 

 tion assigned to ValvuUna, it follows naturally that genera so closely allied should be 

 classed with it amongst the Imperforata. 



There is another phase of this subject which must be alluded to in passing, namely, the 

 relation of the intermediate types, the modifications of which form so important an item 

 in the Carboniferous fauna, to the Rhizopoda of subsequent geological epochs. It is true 

 that a few truly arenaceous species have been met with in Carboniferous rocks, and that 

 all of the three families of the " Perforata" are also represented, though how sparingly, 

 except for the genus Fusulina, we shall presently see ; but the fact still remains, that by 

 far the largest number of Carboniferous Foraminifera, both of species and individuals, 

 belong to genera which under some conditions have arenaceous imperforate tests, and 

 under others are smooth and in some cases perforate. That these should be followed in 

 geological time by one set of isomorphs much more characteristically sandy, and by another 

 set of isomorphs distinctively hyaline and porous, is a very significant fact. Take a single 

 instance the simple uniserial, quasi-Nodosarian type Nodosinella, found in the Carboni- 



1 Prof. Huxley, 'Journ. Linn. Soc. London,' vol. xii ("Zoology"), p. 226. 



