tion has long been current, and I must plead guilty to some participation in statements tending to perpetuate the mistake. It may be easily found, however, that even the maximum pressures noted above would not retard transpiration as much as ten per cent, from that which would take place with a sap of distilled water. One of my reviewers has recently made a variation of this mistake in suggesting that acidity would have a retarding effect on water loss. No foundation exists for such a supposition.
Almost any ordinary branching plant with broad leaves will, if forced to carry out its development under arid conditions, show some of the features of the type of desert plants described, and it is customary to assume that the causal conditions responsible for such forms are the desert factors: that we have here a direct adaptation or environic response which has become heritable in the strictest and fullest sense. This is a matter that deserves the fullest consideration. Meanwhile it will be perfectly safe to assume that such spinose forms represent the simplest or most elementary specializations of desert plants, and species with the most diverse morphological constitution may show alterations of this character. The sclerophylls of the American desert include species of Prosopis, Acacia, Calliandra, Parkinsonia, Cercidium, Olneya of the leguminous plants, Covillea and Zizyphus of the Zygophyllaceæ, Fouqueria, Lycium, Koehberlinia, Condalia, Manzanita, Franseria, Jatropha, Sapindus, Vauquelinia, Quercus, Aster and others.
Southwestern America has been arid for an extremely long period, not uniformly so, however. The researches of Professor Ellsworth Huntington, in which evidence has been obtained from ruins of structures built by man, of geological terraces, lake beds, strands and drainage lines in Central Asia, Palestine, and America, and also by the examination of the structure of the big trees of California, seem to justify the conclusion that variations in climate with regard to temperature and moisture have taken place within the last two thousand years that would be of profound biological importance. It seems fair to assume that similar oscillations, each movement of which might extend over a few hundred years, have taken place previously.
It is under these conditions therefore that we are to think of the evolution of the desert vegetation of the southwest, and present knowledge compels us to believe that much of it originated somewhere within the limits of the region which is arid at the present time. Perhaps the most important constitutents of this indigenous specialized flora are the cacti, which must have originated somewhere in the Mexican highlands in the Tertiary or later. This group is known to contain over a thousand species, and now extends through South America, its distribution offering some most highly localized occurrences of species. So rapid has been its evolution, and so wide the amplitude of its depart-
