181 4. J Imperial Institute, 389 



case they are determined from the base, or the predominating sub- 

 stance; and in both cases the number of constituents, and the 

 structure resulting from tbeir mode of union, serve to distinguish 

 the species. The aggregated rocks are divided according as the 

 cement which unites them is more or less apparent, and according 

 to the nature of that cement, and the grains which occur imbedded 

 in it. 



In tin's undertaking, so important to serve as a base to the history 

 properly so called of rocks, the author lias almost every where pre- 

 served the names given to them by M. Hatty in the arrangement of 

 them which he has made in the Museum of Natural History. 



M. Brogniart has likewise stated to the Class the division which 

 he considers it as necessary to establish among rocks, considered 

 with respect to the epoch of their formation, and to the remains of 

 orgauized bodies which they contain, and which are the best marked 

 characteristics of these epochs. Below all the rest occur the gra- 

 nitic formations destitute of all organic remains, the most ancient 

 that we know. The beds that cover them contain only a very few 

 remains of organized beings, and almost all belonging to the class of 

 zoophytes. A third series, that of the syenitic beds, contains no 

 traces of them ; as if their formation had been for a time inter- 

 rupted. In the fourth series shells begin to appear, especially those 

 which have been called cornua ammonia, and which belong to the 

 family of dry shells. The filth and six series of beds are charac- 

 terized by the gryphitesand cerites, which predominate among their 

 shells. Finally, there are beds in which the distribution is so irre- 

 gular that they cannot be classed according to the order of time. 

 These are the trap rocks on the one side, and on the other those 

 produced by the ejections of volcanoes. In all these groups are 

 mixed transported beds (tern-ins de transport) produced by the 

 violent movements occasioned by successive revolutions, and indi- 

 cating accurately the time when each commenced. 



VEGETABLE PHYSIOLOGY AND BOTANY. 



The fall of leaves in autumn, a phenomenon so well known, has 

 occasioned some discussions respecting its cause, and several obser- 

 vations have been made reapecting the variations in it. M. Carnot, 

 member of the mechanical section of the Institute, but who does 

 not neglect any thing that can furnish a subject for meditation, 

 having remarked that some trees begin to lose their leaves at the 

 top of their branches, and others at the bottom, M. Palisot de 

 I ivois, member of the section of botany, has endeavoured to 

 discover the reason of this difference. He has found that in general 

 autumnal shoots consist of prolongations of the 

 extremities of the branches, begin to lose their leaves at the under 

 . while thosi whose !' consist in small lateral branches. 

 begin to lo-e their leaves at the summits; or in Other terms, that 

 the last haves which expand are likewise the lust that tall oil', 

 Duhamel, who 1 had made a similar remark, was surprised that thoM 



