October 8, 1915] 



SCIENCE 



501 



productivity, or correlates best witli tiie na- 

 tive vegetation. 



In the reports of the Census Bureau there 

 are statistics, made so thoroughly as to prac- 

 tically eliminate the personal equation, that 

 throw valuable light on both evergreen per- 

 centages and soil fertility. The percentage of 

 evergreens is one of the most conspicuous and 

 easily determined features of vegetation that 

 can be expressed numerically, and it is almost 

 the only one (and certainly the most signifi- 

 cant one) that can be calculated from existing 

 census statistics. 



In 1911 the U. S. Census Bureau and Forest 

 Service jointly published an octavo bulletin 

 (without a number) entitled "Forest Prod- 

 ucts of the United States, 1909," giving among 

 other things the output of lumber, laths and 

 shingles by the sawmills of the United States 

 in the year named, by states and by species 

 simultaneously. Similar statistics have been 

 published for subsequent years, but those for 

 1909, which were gathered in connection with 

 the regular decennial census, seem to be the 

 most complete. 



Other things being equal, the lumber sawn 

 from deciduous and evergreen trees in a given 

 area should be proportional to the percentage 

 of those trees in the forests; but other things 

 are not equal. In the first place, it is of 

 course the percentage of evergreens in the 

 primeval forests that counts, and as the de- 

 ciduous trees prefer the richer soils, a larger 

 proportion of them than of the evergreens 

 have been destroyed by farmers in clearing 

 land for cultivation. Second, the evergreens 

 in the United States are mostly (counting in- 

 dividuals, not species) conifers, and conifers 

 on account of their gregarious habit and 

 straight-grained easily worked wood are more 

 sought after by lumbermen than the hard- 

 woods are. 



In the census statistics of lumber produc- 

 tion all the hardwood trees listed are decidu- 

 ous. (Some evergreen hardwoods are cut in 

 the southeastern states, but in insignificant 

 quantities.) The figures for evergreens are 

 therefore obtained by subtracting from the 

 total for conifers those for the two deciduous 



genera, Taxodium and Larix (cypress and 

 tamarack). To make allowance for the hard- 

 woods that have been destroyed by farmers, 

 and the neglect of the remaining ones by 

 lumbermen, the figures for them (not for all 

 deciduous trees) are arbitrarily multiplied by 

 4. This product added to the figures for 

 conifers and divided into the total evergreens 

 (not total conifers) gives a percentage which 

 is believed to approximate pretty closely in 

 most cases the percentage of evergreens in the 

 original forests. 



Some idea of the fertility of the soil in dif- 

 ferent parts of the country can be obtained 

 from several different sets of figures in the 

 census reports. In rural districts there is a 

 close relation between soil fertility and den- 

 sity of population; but in the northeastern 

 states so large a population is supported by 

 manufacturing, independently of the subja- 

 cent soil, that statistics of population there 

 would be of very little use in this connection. 

 Figures showing the value of land are open to 

 the same objection. 



The census gives the acreage of " improved 

 land " in the farms of each state and county, 

 and that is undoubtedly correlated with soil 

 fertility, whether the fertility is due to phys- 

 ical or chemical conditions or something else; 

 and the tendency to utilize the land more fully 

 in the neighborhood of cities is partly offset 

 by the occupation of much land for other pur- 

 poses than farming, and such land is not in- 

 cluded under " improved land in farms." For 

 the purposes of this investigation the statistics 

 of improved land have been taken from the 

 Tenth Census (1880), because earlier censuses 

 are probably less complete, and because the 

 use of commercial fertilizers has increased so 

 much since then as to tend to obliterate differ- 

 ences in productivity between different kinds 

 of soil.^ 



Another kind of statistics given for each 

 state and county in recent census reports is the 



2 In the South some sandy soils which were re- 

 garded as almost worthless half a century ago are 

 valued more highly at the present time than the 

 rich clayey or calcareous soils, because they are 

 more easily tilled, and yield large returns for the 

 amounts invested in fertilizers. 



