57° NATURE [APRIL 14, 1904 
still other constituents—few though they be—like | the desert give rise to self-supporting descendants, for 
Acanthosicyos horrida and Welwitschia mirabilis, are | Ipomoea arborescens is a tree thirty feet in height. Another 
remarkable assemblage of plants is to be seen in the same 
desert at the harbour of Guaymas, in the salt waters of 
which tropical mangrove plants (Avicennia and Rhizophora) 
are growing side by side with typical desert plants, in- 
cluding two species of Cereus. 
Quite apart from a desert climate, waste tracts can be 
induced by physiologically dry substrata, such as rocks, 
shifting sand, gravel, saline or allaline soils, and these 
soils, or high winds, may accentuate the sterility of climatic 
deserts. All these conditions are to be met with within 
relatively easy reach of Tucson, Arizona, the site of the 
laboratory, as is shown by the present report, written by 
Messrs. F. V. Coville and D. T. MacDougal. Desert tracts 
occur varying from the most forbidding and bare areas 
with meagre floras to richer ones like that near Tucson, 
the desert flora of which in an eastward direction gradually 
gives way to the luxuriant subtropical flora of eastern 
Texas. 
The following particulars culled from the report serve to 
illustrate the great variety in the deserts within reasonable 
distance of Tucson laboratory. 
South of Tucson lies the desert of Chihuahua, ‘‘ with a 
long stretch of sand dunes.’’ Here the sand is siliceous, 
and Yucca radiosa, with its immensely long horizontal 
roots, plays a great réle as a sand-binder. 
The Tularosa desert, on the other hand, exhibits a re- 
markable area of white shifting sand, mainly composed of 
calcium sulphate. On these soluble ‘‘ white sands’’ Rhus 
triloba, forming hemispherical bushes, fixes the sand, and 
Populus fremontii lives. 
Lying farther north than Tucson, the Colorado desert 
includes gravel-hills, sand-dunes, alkali-flats, as well as wet 
saline and alkaline spots. Here occur unique groves of a 
fan-leaved palm, Neowashingtonia filifera, growing on 
alkali-encrusted soil above a moist clay subsoil. Over 
portions of this desert ‘‘ the vegetation is subject to a 
veritable sand-blast,’’ which threatens to sever the telegraph 
poles at a height of two feet, and has carved the creosote 
bushes (Covillea) into ‘‘ most fantastic shapes.’’ 
The report includes a record of a few preliminary observ- 
ations made by Dr. MacDougal, and a bibliography com- 
: ~ piled by Dr. W. A. Cannon, the resident investigator. Its 
TT prey UALS ac eat ad saeanrens) near Tucson, Arizona. aie and interest are enhanced by twenty-nine excellent 
: i a photographs of vegetation in the deserts. 
enigmatic and isolated species the presence of which defies | For the establishment of this laboratory botany owes a 
explanation. On the geographical distribution of species in | debt not only to the munificent founder, Mr. Carnegie, 
deserts questions present themselves in 
numbers. For instance, why are certain 7 
species limited to specific areas of a 
large continuous desert? Competition 
among the component plants is practi- 
cally eliminated, save, perhaps, in the 
seedling stage. Is the restriction of 
area a consequence of temperature, and 
therefore indirectly one of lineage, or 
of the chemical or physical nature of the 
soil, or of a number of factors? The 
vast North American desert district 
affords admirable opportunity for 
answering these and other questions, as 
it stretches, with small interruptions, 
from Mexico northward to the heart of 
the United States, and its different areas 
display differences in their floras. The 
Mexican deserts, for instance, give 
blurred impressions of an_ enfeebled 
tropical flora, inasmuch as they include 
Hamatoxylon and Guaiacum, which are 
absent farther north. The Sonora 
desert in Mexico shows a somewhat ex- 
ceptional feature in the marked presence 
of climbers, including malpighiaceous 
lianes and two other remarkable plants, 
a tuber-rooted straggling Cereus and - ; 
the tendrilled cucurbitaceous I[bervillea Fic. 2.—Group of Palms (Weowashingtonia filifera) in the Colorado desert, California. 
sonorae, which in form recalls Testu- 
dinaria. Yet in this same desert’ there is not want- | but also to the suggester of the scheme, Mr. F. Vv. 
ing evidence in favour of the view that climbers driven into | Coville. Percy GROOM. 
NO. 1798, VOL. 69} 
