Mr. H. J. Carter on the Fossil Foraminifera of Scinde. 419 



are reached, they undergo the change common to the discoid 

 Foraminifera; that is to say, their vertical axis becomes the 

 longest, from the diminution of the horizontal one (i). The 

 chambers also vary greatly in length, and this causes a corre- 

 sponding irregularity in the rows of which they form a part, — 

 thus reducing the row gradually almost to a mere line, and then 

 expanding it out again to its full size (PI. XVII. fig. 1 n) ; at 

 other times the chamber becomes doubled, and thus the row 

 appears to bifurcate as just stated; while on other occasions, 

 again, the opposite takes place, viz. the chamber ceases to be 

 developed, and the adjoining rows closing in, the abortive row 

 thus terminates or ceases to exist. The additional row or bi- 

 furcation probably begins in an offset from one of the annular 

 canals which will be presently mentioned, and should be regarded 

 as a "branch," like the branches of the spire in Nummulites. 

 (Reference to the figures in the first row of PI. XVII. will facili- 

 tate the explanation of all this.) 



Lastly, the chambers are, for the most part, arranged alter- 

 nately in adjoining rows, and each communicates by two canals 

 or stolons with the two immediately before and behind it ; so 

 that every chamber has four others in connexion with it (/c, /). 

 In this we have an arrangement analogous to the "oblique" 

 canal-system in Qrbitolites Mantelli, which will be described 

 under this head. 



Canal-system. — Here and there my horizontal sections present 

 a distinct canal-system (&3, I3), consisting of single annular 

 canals situated between the rows of chambers respectively, con- 

 nected with each other by straight smaller branches which pass 

 through the interseptal spaces; so that each chamber is thus 

 enclosed in a quadrangular mesh of canals, and the whole toge- 

 ther form a meshwork plane which is double ; that is to say, one 

 exists on each side the horizontal plane of chambers, on a level 

 with the chambers, so as to have the stolon-processes between 

 them. Why the canal-system should only appear here and there 

 in remnants in my sections, I cannot say, unless it be from 

 variation in their size, from imperfect infiltration, or from total 

 absence in parts. Again, from their being analogous to the annular 

 canals of Orbitolites Mantelli, which will be described presently, 

 as well as to the great spiral canals of Operculina and Nummulites, 

 one would expect the proximal ends of the chambers to be united 

 to them : and this is actually the case ; that is, the chambers 

 have often a bond of union of this kind when the entirety of the 

 annular canal is not visible, and are as often without it when the 

 annular canal, in its entirety and separation, is present (k) ; while 

 I know of no other structure of the kind in the horizontal plane 

 but this annular canal-system and the stolon-process, except, 



Ann. $ Mag. N, Hist. Ser. 3. Vol. viii. 29 



