598 JOURNAL OF FORESTRY 



is often as great in extent as that put to economic use. In the aggre- 

 gate the right of way for country highways occupies a vast acreage, 

 of which the actual roadbed is but a small part. In most European 

 countries the land is utilized to the very edge of the roadbed, but in this 

 country there is a strip of unused land at either side extending to the 

 boundary of the right of way. These strips of unused land are com- 

 monly about one and a half rods wide and require considerable atten- 

 tion on the part of highway commissioners and local tree wardens tr- 

 keep them free from weeds and brush. Periodic "brushing out" of 

 these strips of idle land can only be done at considerable cost, which 

 the author considers a waste of money, as the improvement is only 

 temporar}-. He believes that forest trees should be planted or encour- 

 aged from natural regeneration, as this silvicultural treatment is the 

 only practical way to keep the right of way free from woods and brush. 

 The trees should be free of lower branches in order to attain air circu- 

 lation across the roadbed and permit an unobstructed view on either 

 side over the adjacent country. The roadbed is shaded in summer and 

 less dusty than one exposed to full sunlight. The snow is prevented 

 from drifting over the roadbed in winter, and when the crowns are 

 sufficiently high and without lower branches the roadbed is overwet 

 only on exceptional sites. 



The author describes a practical method for developing roadside for- 

 ests and gives lists of species most useful on different sites in New 

 England. The bulletin is illustrated by photographs showing roadbeds 

 bordered by weeds and brush which require frequent attention and 

 others bordered by stands of timber of suitable character to keep the 

 roadbed in good condition. 



J. W. T. 



I'ariatioji in the Chemical Composition of Soils. V>y W. O. Robin- 

 son, L. A. Steinkoenig, and William H. Fry. U. S. Department of 

 Agriculture, Bureau of Soils. P>ulletin 551. Washington, D. C. 1917. 



The purpose of this paper is to show the magnitude of variation in 

 the chemical composition of American soils. The data are from a 

 previous paper giving the complete analysis of a number of important 

 American soils supplemented by the analysis of 45 samples represent- 

 ing 18 distinct soil types given for the first time in the present paper. 



The paper discusses the preparation of samples and gives the meth- 

 ods and results of analysis. There is also presented a discussion of the 

 data in which special consideration is gi\en to the relation of soil to 



