From this it is to be seen that the localities referred to, instead of being typically intense deserts, are, on the other hand, the most favorable situations as regards moisture.
It is difficult, at present, to state under what conditions the roots of the desert plants are formed, owing mainly to the lack of experimental evidence. But by a system of reasoning backwards from the mature root we can possibly picture to ourselves something of these conditions. In the first place, if we examine the root-systems of desert plants, in the field, during the season of drought, we shall find it very difficult, if not impossible, to find any portions which show vegetative activity, although it may be possible at the same time to demonstrate a certain, even if low, rate of transpiration. On the other hand, if the root-systems of the desert plants are examined during the rainy periods, there will be no difficulty whatever in finding fresh growth, new rootlets of whatever kind. But that this is not the whole story is evidenced by the fact that in winter many of the plants native to the southwest do not form new roots, or, at least, I have not been aide to find new roots. In spite of this fact, such plants as the flat opuntias do, in winter, absorb water and very promptly after rains. This is shown by the thickening of the fleshy and flat stems. It is therefore probable that a certain amount of heat as well as of moisture is required to bring about the formation of fresh roots. In addition to these two factors, there is probably another one, namely, aeration of the soil. Whether this is mainly concerned with the formation of the roots or of the position occupied by the roots in the soil is not known. It seems highly probable in certain cases, particularly in fleshy plants like the cacti and some liliaceous forms, that the amount of air in the soil must be of importance in determining the position occupied by the roots. So far as observation goes, the roots formed may be classed in at least two categories: (1) They constitute the extension of the roots previously formed and (2) they may appear on much older roots, but are of limited growth. It is supposed that in the main the greatest amount of water taken into the plant comes through the roots of the first kind, so that the place of water absorption as the roots grow, ever becomes farther from the stem, and the problem of water transportation is ever an increasingly difficult one. This last one is probably to be considered a very important matter on the desert where the evaporation rate is often very high, caused by the low relative humidity, by high temperature and by air currents. It is conceivable that, given favorable conditions, a large proportion, possibly all, of the roots of this character might remain alive, but, as a matter of fact, in desert plants, as before noted, it is difficult during the dry seasons to find any living roots of this class. As one result of this we find that the extension of the root-systems as a whole, away from the central plant axis, goes on relatively slowly, and
