GENERAL SUMMARY AND CONCLUSIONS. 67 



As is well known, a desert flora is in part perennial, lasting with but little 

 outward change from season to season, and in part ephemeral, consisting 

 of short-lived species which appear with the rains and which disappear 

 with the return of the dry season. The ephemeral flora differs in no essen- 

 tial regard from annuals of the moister regions; also, the environment to 

 which they are exposed closely resembles the environment of the annuals 

 of such regions. But the perennial desert flora, on the other hand, offers 

 very striking departures from the corresponding flora of the moister regions, 

 just as the environment to which they are exposed for the most of the year 

 is also different. It will be sufficient, for the purpose of bringing out the 

 point of view, to notice a few of the leading characteristics of desert plants 

 and of their physical environment and, in a few instances, to observe pos- 

 sible relations between the two. 



The most obvious features of desert plants are associated, in whatever 

 way, with the subaerial portions. Leaves are usually greatly reduced or 

 wanting, during the dry seasons at least. Spines are frequently present 

 and the exposed parts are often well covered with hairs. The stomata 

 are sometimes deeply sunken, the cuticle often very heavy, and a waxy 

 substance may cover leaves or stems. The chlorophyll-bearing cells are 

 arranged with the long axis at right angles to the leaf or stem surface. All 

 or most of these characters are associated with the low humidity of the 

 air. In certain deserts plants are also to be found with greatly enlarged 

 stems and branches which serve as water-storage organs. It should be 

 noted, however, that plants which do not have a constant surface undergo 

 many marked changes with a betterment of the water relations, partictdarly 

 if this comes when the temperatures are favorable. For example, many 

 cacti organize leaves which are unsuited in structure for periods of extreme 

 drought, and which consequently fall away soon after the close of the rainy 

 season. These leaves enormously increase the rate of transpiration at a 

 time when this is not injurious. 



It is not in the subaerial parts alone, however, that the plants of one 

 desert are different from those of another, that plants are unlike in the 

 same desert, or that plants of a desert are different from those of the more 

 humid regions. The root-habits also exhibit not a little diversity and show 

 marked reactions to the pressure of their en\'ironment. For instance, the 

 desert shrubs of the region surrounding the Desert Laboratory at Tucson 

 have well-marked root-systems, apparently constant imder natiu-al condi- 

 tions, which may roughly be designated as the tap-root type, the super- 

 ficial type, and the generalized type. Other conditions being equal, species 

 with characteristic root-types have also characteristic distribution, or 

 exhibit in other regards consistent reactions. Thus, the widely extending 

 and superficial type of roots is confined, among independent plants, to 

 such as have water-storage capacity. The relation of this root-type to the 

 distribution will be mentioned later. Plants vnth a dominating tap-root 



