January 24, 1908] 



SCIENCE 



141 



neyed gathering-up of superficial observations 

 on " plant associations," without any mention 

 of the probable, in many cases abundantly 

 obvious, causes of the geographical grouping 

 of plants. Ecological studies, as often made 

 savor strongly of the " gedankenlose Heusamm- 

 ler " habit animadverted upon by Schleiden 

 over half a century ago, and were apparently 

 only temporarily stopped by Darwin's great 

 work. The soil-conditions accompanying the 

 occurrence of certain plant groupings are 

 usually so superficially set forth that nothing 

 but the old classification into hydrophytes, 

 mesophytes and xerophytes is attempted; in 

 conformity with a hypothesis based upon the 

 arbitrary assumption that moisture is the only 

 controlling factor of plant growth. Adding 

 to this hypothesis the factor of soil-texture, 

 and basing thereon the entire work of soil 

 classification, Whitney and the Bureau of 

 Soils of the United States have built up a one- 

 sided theory, which is in fiagrant contradiction 

 to facts observable by any one not under the 

 oiEeial afflatus of that head center. Some 

 years ago E. M. Harper, in his studies of the 

 plant geography of the southern Coastal Plain, 

 entered upon the right track so far as observa- 

 tions in the field are concerned; but Fernald 

 has added to the field observations a closer 

 discussion of the exact physico-chemical fac- 

 tors which condition plant distribution, such 



, as I have urged for fifty years past. 



In defining his investigations as relating to 

 certain " alpine " plants, Fernald really blurs 

 the nature and importance of his work. Alps 

 are usually rocky and have predominantly 

 sedentary soils, that is, soils overlying the 

 rocks from which they have been formed by 

 weathering, without having received admix- 

 tures of the decomposition-products of other 

 rocks. Such admixtures may, nevertheless, 

 easily occur locally. Moreover, the fact is that 

 the flora of mountains is often largely paludal 

 on account of their " young " geological sur- 

 face features, and therefore commonly include 

 wet meadows, ponds and lakelets alongside of 



- cliffs and rocky slopes. The frequent wash- 

 ing-down of the decomposition-products of 

 various rocks on the higher slopes to those 

 lower dovm, has made the word " alpine flora " 



a designation referring mainly to temperature- 

 conditions. Fernald himself, however, at once 

 recognizes the distribution of the " subalpine " 

 flora to be almost identical with that of a large 

 portion of the lowlands. 



Fernald presents (pp. 158 to 164) lists of 

 258 plants belonging to the territories of New 

 England and Canada, tabulated so as to show 

 their occurrence, either preferably or exclu- 

 sively, upon rocks considered respectively as 

 potassie,^ calcareous and magnesian. 



These tables show, in three columns, the 

 plants belonging to soils formed from (sup- 

 posedly) potassic, calcareous and magnesian 

 rocks, and exhibit the very striking contrasts, 

 as well as some coincidences, in the floras in- 

 habiting the three classes of rocks. 



Among the most striking of these is the 

 often-observed poverty of the vegetation on 

 soils derived from exclusively or chiefly mag- 

 nesian rocks, such as serpentine and talc 

 schist ; to which may be added dolomite, which, 

 though consisting more than half of lime 

 carbonate, shows everywhere a more or less de- 

 pauperated flora, especially when compared 

 with adjacent areas of non-magnesian lime- 

 stones. Eernald shows how, on the north 

 slope of Mt. Albert, a serpentinous plateau in 

 Gaspe county, by the partial admixture of cal- 

 careous and potassic rocks an immediate in- 

 crease of species and a more luxuriant de- 

 velopment take place. A very few plants 

 only are known to prefer purely magnesian 

 soils, everywhere. Among these, Fernald 

 mentions specially, Cerastium arvense, whose 

 ash contains the unusual proportion of 19.8 

 per cent, of MgO. 



Fernald promises, by analyses of the ashes 

 of a number of these plants (as well as, it is to 

 be hoped, of corresponding soils) to establish 

 definitely the relations between their plant-food 

 requirements and their rock habitats. As re- 

 gards the soils, it is evidently of high impor- 

 tance that the analyses should be made by such 

 methods as will show their permanent char- 

 acter, and not merely the accidental present 



' Wlietlier the granites, gneisses and mica schists 

 assumed by Fernald as producing soils rich in 

 potash really do so, is somewhat questionable, and 

 should be verified by analysis. 



