394 



SCIENCE 



[N. S. Vol. LI. No. 1320 



rapid drying out of the surface sand a dust 

 mulch is formed which retards effectively 

 further water loss from the soil. At a depth 

 of a few inches the soil is always moist, and, 

 from data given for another locality with 

 similar soil, it would appear that the moisture 

 may be fairly uniform to a depth of six feet. 

 Exact data, however, as regards this feature 

 are wanting. Of 19 sandhill species whose 

 roots were studied, 8 have roots which are 

 entirely or nearly confined to the first two 

 feet of soil, and of the balance all save one 

 have the greatest root development at this 

 depth. The roots of Ipomeea were the most 

 extensive of those of any species in the com- 

 munity, or, for that matter, apparently the 

 most extensive of any observed during the 

 course of the study. The block of soil in- 

 cluded within their reach was approximately 

 fifty feet in diameter and over ten feet in 

 depth. The roots were fairly well distributed 

 throughout except only in the surface foot 

 from which they were largely wanting. An- 

 other feature of the root was the enlarged 

 and tapering tap which was about eight inches 

 in diameter a foot beneath the surface and 

 the enlarged portion of which was about three 

 feet long. The enlarged tap of Ipomosa con- 

 stitutes an important reservoir for food and 

 water storage. 



Weaver finds in general that in the com- 

 munities studied the most striking root char- 

 acters, at least so far as the gross morphology 

 is concerned, are intimately related to the 

 moisture conditions of the soil. Where, for 

 example, the uppermost soil layers only are 

 moist, there is a marked development of 

 laterals. In the event the soil carries mois- 

 ture to a considerable depth, as on the 

 prairies, deep root penetration in many spe- 

 cies occurs. Apparently he does not find soil 

 temperatures or soil aeration limiting factors 

 in root penetration although that such may 

 be the case in certain instances seems to the 

 reviewer not unlikely. For example, the 

 roots of Opuntia fragilis do not appear to 

 attain to a depth greater than fifteen inches, 

 and it is usually considerably less than this. 

 The roots of Yucca are also for the most part 



shallowly placed. And, finally, in the prairies 

 as regards penetration, there is a fairly well- 

 marked stratification of the roots. It may be 

 as suggested in the case of plains species that 

 the " well developed system of shallow, widely 

 spreading laterals is undoubtedly a response 

 to the moisture in the surface soils resulting 

 from frequent light smnmer showers." How- 

 ever, in the opinion of the reviewer, the pos- 

 sibility that the root-temperature or the root- 

 soil aeration relation may also be of im- 

 portance is by no means excluded. The 

 various root relations are so closely inter- 

 woven that any one can only be evaluated 

 when the rest are so far as possible con- 

 trolled. And this requires exhaustive experi- 

 mentation, which was not within the scope 

 of the present study. 



The extremes as regards root penetration 

 appears to be met in the case of Opuntia 

 fragilis, of the plains, on the one hand, and 

 possibly, Lygodesmia juncea, of the iffebraska 

 prairies. In Opuntia most of the roots lie 

 within one to three inches of the surface of 

 the ground, with an extreme penetration of 

 eight to fifteen inches. While the roots of 

 Lygodesmia have been found to attain a 

 depth exceeding twenty feet seven inches. In 

 the latter instance the soil is loess, with 

 uniform physical properties, and is very 

 favorably for deep root penetration. This 

 well authenticated penetration is sufficiently 

 deep, but it is of interest to note the observa- 

 tion given in MerilP that " Aughey has 

 found roots of the buffalo berry (Shepherdia 

 argophylla) penetrating the loess soils of 

 ISTebraska to a depth of fifty feet." 



In a work so well done it seems captious to 

 allude to a feature not by itself of funda- 

 mental importance. However that may be, 

 it seems to the reviewer unfortunate that the 

 English and the metric systems of measure- 

 ment, especially, are both used throughout 

 the study. Consistency in this regard would 

 surely meet more general approval. 



W. A. Cannon 

 Desert Labokatort 

 2 "Rocks and Eock Weathermg, " p. 181. 



