GEOGRAPHY OF PLANTS. 49 



moisture, and exposure to certain winds. Of the forty-eight species enu- 

 merated as having a north and south distribution, only two are natives of 

 Britain, namely, Pinguicula lusitanica, which extends along the Atlantic 

 shores from Portugal to the north of Scotland, but which is not found any- 

 where far from a western coast ; and Erythrcea latifolia, likewise a coast 

 species, found from Norway to Portugal. 



The distribution of individuals in the areas occupied by the species is 

 next discussed, with considerable detail, showing how local causes modify 

 the frequency or non-frequency, or the luxuriant development, of each 

 species. The nature of the soil, exposure, supply of moisture, and other 

 obvious modifying causes are indicated, and a list is given, after Mohl, of 

 species which are characteristic of primitive rocks, and of those which only 

 occur in calcareous soil. M. de Candolle seems to think also that some- 

 thing like a " rotation of crops" exists naturally among wild plants, and 

 that a species shifts its soil (especially an annual one) from year to year, 

 from causes similar to those that force the farmer to vary his seeds when 

 cultivating the same soil. " One cannot doubt," he says, " that the exist- 

 ence of a species, and especially its prolonged existence, becomes a cause 

 unfavourable to the life of that same species, or of analogous species, in 

 the same soil." The well-known fact that when a natural forest is burned 

 or cut down, trees of different species commonly spring up in the room of 

 those destroyed is adduced in proof of the necessity of rotation ; and the 

 hypothesis of the elder De Candolle, that roots discharge excretory matter, 

 so as to poison the soil for themselves, is dwelt upon as an established fact. 



The seventh chapter treats of the area or space on the surface of the 

 globe over which a species extends. This important subject has for a long 

 time engaged the author's attention, and the result now given must have 

 cost years of toilsome research. The value of this research must, however, 

 in great measure depend on the author's views as to the limits of the 

 species themselves. In the hands of some writers who establish " new 

 species" on every local race, if it differ by a hair's breadth from their 

 " typical" form, a chapter like the present would only lead to confusion. 

 Fortunately for science, M. Alphonse de Candolle is content to call " Ra- 

 nunculus aquatilis" by its single name, and consequently finds it to extend 

 over the northern hemisphere, from Lapland to Abyssinia, and from the 

 68° parallel of north lat. in America to California. Had he chosen the 

 opposite course, he might easily have treated us to a crowd of " representa- 

 tive species," each peculiar to its own pond or ditch, over the same extent 

 of surface. 



