Dr. W. B. Carpenter on the Polyti-emata. 383 



grouped in accordance with the direction of their modification. 

 Thus, taking the Milioline series as an illustration, we accept 

 Spiroloculinaj Biloculina, Triloculina^ and QuinquelocuUna^ 

 not (in the sense of D'Orbigny) as generic names of groups 

 capable of being sharply differentiated from each other, but 

 as designations of certain well-marked types that may be 

 conveniently adopted as points of departure for the orderly 

 arrangement of those multitudinous specific and varietal modifi- 

 cations which, when thus studied, are found to constitute a 

 continuous nexus that defies all attempts at classification by 

 strict definition. So, I should suppose, no one would think 

 of abolishing generic types so strikingly differentiated as 

 Cornuspira and Orhitolites because both of them in their 

 earliest stage of growth often correspond with the Milioline 

 Spiroloculina. Nor should we be wise in abandoning the 

 generic distinction between Orhitolites and Orhiculina be- 

 cause, in the later stages of their growth, marginal fragments of 

 the disks of these two types cannot be distinguished from 

 each other. Nor, again, does the discovery by M. Munier- 

 Chalmas of a type most curiously intermediate between 

 PeneropUs and the spiral Orhiculina (the continuous chambers 

 of the former being partly subdivided by transverse inden- 

 tations of their walls, so as to take the form of moniliform 

 rows of freely communicating chamberlets) invalidate the 

 propriety of retaining those two well-characterized types as 

 generic centres. The same is preeminently true of the Oris- 

 tellarian and Rotalian groups, and still more, if possible, of 

 those Arenaceous forms, often bearing a most curious iso- 

 morphic resemblance to the calcareous-shelled Foraminifera, 

 which are among the most remarkable novelties brought to 

 light by recent Deep-Sea explorations. In fact, if we say 

 that in each of the principal series of Foraminifera " every 

 thing graduates into every thing else," we shall not be far from 

 the truth. 



If, then, we agree to retain as generic centres the forms most 

 strongly differentiated in their plans of growth, I maintain 

 that the typical Carpenteria is generically distinct from the 

 typical Polytrema. The latter, as Mr. Carter truly says, is 

 essentially a hranching stiaicture ; and the base from which it 

 rises, in all the instances in which I have examined it, has 

 (like the primordial })lane of Tinoporus) more or less of the 

 " Planorbuline " arrangement, the Rotaline spiral very early 

 giving place to the cyclical mode of increase. The upward 

 growth of this branching structure essentially consists in a 

 vertical piling-up of minute chaml)ers resembling those of the 

 basal disk ; and the distinctive peculiarity of the typical Poly- 



