344 Proceedings of the 



which shows us at each change a new order of things doubtless in 

 harmony with the subsequent atmospheric circumstances. The im- 

 portance of these considerations, when united with the observations 

 recently made on the fossil remains of animals, and the researches 

 of M. Elie de Beaumont on the relative ages of the mountains of 

 crystallization, will be obvious to every geologist. M. Brogniart has 

 also gone into some curious researches on the comparison of the 

 laws regulating the different epochs of vegetation in the ancient 

 world with those now existing. The vegetable remains of the first 

 period have much greater resemblance to those now found near the 

 equator than to those of the temperate zone. In the torrid zone 

 we find much higher species of horse-tails, ferns, and lycopodia, 

 than are observed near the poles : many of them are even arbo- 

 rescent there. But the size of the fossil plants of those families in 

 the first period of vegetation is much greater than even that of 

 those now existing at the equator, consequently the causes of de- 

 velopment were then much more intense than at present. On the 

 other hand, M. Brogniart remarks that these vascular cryptogamiae 

 formed T 9 ^ths of the primordial vegetation, whereas they scarcely 

 constitute -f^th f ^ e ac tual vegetation ; there must, therefore, have 

 been some cause producing this preponderance. It had been before 

 remarked that these cryptogamise attain the greatest size between the 

 tropics ; but M. Brogniart is the first to call attention to the fact, 

 that their relative quantity is greatest in islands, and becomes more 

 considerable in proportion as the islands are smaller and at a dis- 

 tance from the continents, so much so that in some small islands 

 they are even now almost equal in number to the phanerogamiee : 

 whence we may conclude, that if the world were composed, not of 

 continents, but of small islands only, the cryptogamise would be 

 equal or superior in number to the phanerogamic. This observa- 

 tion is of great importance when considered in conjunction with the 

 geological facts which induce us to conclude that the existing conti- 

 nents were not always in their present state, but have been formed 

 piecemeal, and that probably, in the most remote ages, the earth con- 

 sisted only of islands and archipelagos in the midst of a vast sea. 

 M. Brogniart then shows us, that, in the subsequent periods, vegeta- 

 tion began to resemble that which we now see on the coasts and in 

 the large equatorial islands, both in the size of the plants, the families 

 to which they belong, and the numerical relation between these fami- 

 lies. Finally, he shows that, in the most recent period after the ap- 

 pearance of the dicotyledonous plants, the vegetation became analo- 

 gous to that now existing in large continents. These observations are 

 derived from the general examination of all M. Brogniart's printed 

 and manuscript labours. The work in course of publication, entitled 

 'L'Histoire des Veg'etaux Fossiles,' will form a complete table of all 

 the fossil plants considered with a view to their classification, and their 

 analogy with the plants now existing. The five numbers which have 

 Appeared, treat of the conferva?, the fuel, the equisetse, the musci or 



