Group of Russian Fusulinae. 421 



habiting estuaries and shallow water, perfectly simple in 

 shell-structure, and in reality little other than a "starved" 

 modification of Rotalia Beccarii. In the littoral and Lami- 

 narian zones of our northern seas the central or typical form 

 [R. Beccaviij Linne) is common ; but the specimens are small 

 and seldom show any advance in organization on the estua- 

 rine variety beyond the occasional duplication of the septal 

 walls. The larger outspread variety which occurs in the 

 Adriatic {R. ammoniformis, D'Orb) exhibits tliis same ten- 

 dency in perhaps a greater degree ; and occasionally both 

 single and double septal lamellee may be seen in the same 

 shell. In the fine, thick, externally granulate examples of 

 the type met with in the West Indies, in the Levant, and 

 elsewhere, distinguished by D'Orbigny as R. corallinarum 

 (Modele no. 84), not only is this duplication a constant cha- 

 racter, but it is accompanied by considerable development 

 of supplementary skeleton and a rudimentary system of inter- 

 septal passages. Lastly, in R. Schrmteriana^ P. & J., the 

 supplementary skeleton and the complicated canal-system 

 become salient features, and denote the highest organization 

 attained by the liotalian type. 



It is clear, therefore, that, from a morphological point of 

 view, the canal-system is to be regarded chiefly as evidence 

 that a certain stage of development in a closely connected 

 series has been reached, and that, however valuable as af- 

 fording collateral characters, it is not available as a basis of 

 distinction amongst genera, still less in the construction of 

 families or other groups of higher relative value. 



Turning again to the Nummulinida, similar series may 

 easily be found — such, for instance, as that commencing Avith 

 the simple brackish-water Nonioi:iina3 and culminating in the 

 highly organized PoJystomella craticidata — a chain in which 

 the successive links are so similar to those of Rotalia that 

 they may be placed in rank, side by side, as isomorphs. 



The bearing of these facts upon tlie position to be assigned 

 te the genus Fusulina is manifest. In their presence the 

 chief argument for its association with the Rotalines falls to 

 the ground, and the only plea for its separation from the 

 Nummulines becomes untenable. On the whole, the sub- 

 genus Nonionina yields the best key to the position and 

 characters of Fusulina — a fact recognized by D'Orbigny 

 thirty years ago, when he placed the latter genus between 

 Nonionina and Nummulina. In modern systems, natural 

 affinity has been more sought as a basis of classification 

 than the artificial distinctions which served so good a pur- 

 pose in the hands of the earlier naturalists ; but in the pre- 



