REVIEWS 791 



of previously felled trees on the i^oor and sides of a small flat valley, 

 through which ran a small stream, dry during a part of the summer. 

 It is to he noted that the trees before felling were isolated in open 

 stand, growing above a sparse undergrowth of shrubs or of grass. 



Two methods were followed in making the measurements. In one a 

 sheet of white paper was laid on the stump and marks thereon indi- 

 cating the limits of each year's growth were made along the edge of 

 the paper. Three such records were obtained from different radii on 

 each stump. The sheets were later taken to the laboratory and the data 

 worked up at convenience. In the other method followed, a ruler was 

 laid on the stump along a given radius and the width of each ring noted 

 and recorded in the field. 



All the stumps measured were of trees that were approximately 100 

 years of age. Detailed studies were made on sixteen trees from which 

 3,500 measurements of individual rings were obtained. The data are 

 presented in tabular form showing the average width of the rings for 

 periods of years from 1830 to 1910. This table appears to show that 

 there was an* increase in the width of the annual rings with increase 

 in the age of the trees up to about 100 years for the trees measured. 

 Thus from 1830 to 1859 the average width of the annual rings was 

 0.052 inch; from 1860 to 1889, 0.106 inch; and from 1890 to 1919, 

 0.152 inch. 



Although the author states that the growth in diameter of a given 

 tree is the resultant of many factors, both internal and external, he is 

 unable, from his data, to explain the increased diameter growth in the 

 trees studied in terms of these factors separately considered. 



The most striking fact brought out from the study of the data is the 

 frequent occurrence of narrow rings alternating with one or several 

 much broader ones. A table is presented which shows the following 

 to have been years of relatively small growth: 1833, 1834, 1838, 1840, 

 1841, 1848, 1860, 1870, 1874, 1877, 1886, 1887, 1888, 1890, 1894, 1900, 

 1901, 1903, 1906, 1911, 1914, and 1918. The attempt by the author to 

 correlate variations in the width of the annual rings with the length of 

 the growing season, the dates of the last killing frost in the spring or 

 the first killing frost in the autumn, met with little success. A com- 

 parison, however, of mean monthly .temperatures with the diameter 

 growth seemed to show that the temperatures of May and June vary 

 inversely with the width of the growth rings. So also a comparison of 

 precipitation with diameter growth show that in some years a drop in 



