838 



SCIENCE. 



[N. S. Vol. XXII. No. 573. 



of Garrya rigida and G. fremontii, and a host 

 of others. 



The genus Rhamnus, as represented on the 

 Pacific slope, offers some excellent illustra- 

 tions. Rhamnus californica of the coast 

 ranges of central California has at least two 

 very near relatives occurring in adjoining 

 regions. R. purshiana of the northwest re- 

 gion and R. tomentella of the foothills of the 

 Sierra Nevada and southern California, And, 

 as we would naturally suspect, from the theory 

 of isolation, the species occurring in central 

 California is the intermediate one. 



Oeanotlius integerrimus (0. andersoni) of 

 the Santa Cruz Mountains, C. nevadensis of 

 the Sierra Nevada and G. puberulus of the 

 San Bernardino Mountains are three very 

 closely related species occurring in three dif- 

 ferent mountain ranges of the California 

 province. 



Adenostegia rigida of the coast ranges of 

 central California, A. filifolia of southern 

 California, and the Sierra Nevada form, not 

 yet clearly defined, but bearing the name A. 

 rigida hrevihracteata, are, also, three very 

 closely related species, clearly marked in the 

 more isolated portions of their ranges, but 

 apparently intergrading where the three 

 ranges converge. 



In southern California may be found other 

 illustrations fully as conclusive. The flora of 

 this region is naturally very similar to the 

 more typical Californian flora, but it also has 

 certain affinities with that of the Arizona 

 region. Here occurs Quercus engelmanni, and 

 it is in the Arizona region that we find its 

 closest relative, Q. oUongifolia. Again, Eu- 

 pliorhia palmeri is represented in Arizona by 

 E. palmeri peplifolia and Geanothus palmeri 

 by G. myrianthus. 



The coastal and desert regions of southern 

 California also present some well-marked ex- 

 amples. In this connection we need only 

 suggest Eriogonum fasciculatum and E. poli- 

 folium, Stenotus linearifolius and 8. interior, 

 Behhia juncea and B. aspera. 



On the mesas about San Diego is the very 

 common shrub, Adenostoma fasciculatum ol- 

 tusifolium, which is wholly replaced north- 

 ward by the typical form. Again Galochortus 



weedii, also limited to the same general re- 

 gion, is replaced by G. weedii purpurascens in 

 the vicinity of Los Angeles and Santa Bar- 

 bara. 



In the above illustrations it will be noted 

 that the species selected are very closely re- 

 lated. Some may be inclined to criticize this 

 and it may be argued that the plants mentioned 

 are, at least in some cases, not distinct species. 

 This we are perfectly willing to admit as 

 plausible. They may be only subspecies ; but 

 they are, nevertheless, just as suggestive of 

 the isolation theory. 



We do not wish it understood, however, that 

 we consider isolation the direct cause of the 

 origin of species; but, whatever the cause, we 

 do maintain that the evidence in favor of iso- 

 lation as an important factor in the perpetua- 

 tion of closely related species is almost over- 

 whelming in plants as well as in animals. 

 And any theory of evolution which will not 

 allow for this fact can not possibly prevail. 



LeRoy Abrams. 



United States National Museum. 



ground rock for fertilizing purposes. 



To THE Editor of Science: For several 

 years the Division of Tests which is now at- 

 tached to the Office of Public Roads has been 

 investigating, in connection with the study of 

 macadam road materials, the cause of the 

 binding power of rock dust. These investi- 

 gations have led to the conclusion that the 

 decompositions that take place in rock powders 

 under the action of water, when in a very fine 

 state of subdivision (180-mesh and finer), and 

 especially when the grinding has been done 

 wet, bear upon a great many practical prob- 

 lems, some of which are of the very highest 

 importance. 



It appeared in fact that many of the feld- 

 spathic rocks which are more or less rich in 

 potash might be made directly available as a 

 fertilizing material. Although somewhat out 

 of the line of experimentation of this office, 

 under proper authorization the writer con- 

 ducted a series of experiments on tobacco 

 seedlings which showed that fine-ground ortho- 

 clase was very nearly, if not quite, as efficient 

 as a source of potash plant food as the more 



