212 EXPEEIMENT STATION EECOED. 



year ended March 31, 1914, are described and tables showing rainfall at various 

 places in the country are given. It is stated that during the period named " the 

 weather over the Union as a whole was characterized by practically the same 

 features as during the preceding twelve months, viz, a shortage of rainfall, 

 unusual warmth, hot, drying winds during spring, with unseasonable late frosts 

 in the east and center of the Cape Province." 



Relative reliability of long-time rainfall observations, F. H. Millaed 

 (Engiii. News, 73 {1915), No. 25, p. 1212, figs. 2).— Studies of a 63-year rainfall 

 record at Milwaukee, Wis., and of a lOO-j-ear record at New Bedford, Mass., 

 to observe the maximum and minimum average rainfall per year obtained by 

 combining consecutive years into groups of varying lengths, brought out the 

 fact that "the variations above the average are considerably larger than the 

 variations below the average, and that if one is to observe for a short period 

 only, the variation above the average is likely to be appreciably larger than 

 the variation below the average. ... In one case the maximum single year 

 exceeded the average by over 60 per cent, while the minimum year was less 

 than the average by 40 per cent. In the other case the figures show a variation 

 of 40 per cent above the average for the maximum and 26 per cent below the 

 average for the minimum. Taking 10-year periods, the maximum isi only 16 

 per cent greater than the average, while the minimum is 13 per cent less in 

 one case, the maximum being 17 per cent greater and the minimum 7 per cent 

 less in the other case." 



Average rainfall in the light of the New Bedford record, N. M. Stineman 

 (Engin. News, 73 (1915), No. 25, p. 1213, fig. i).— Studies of a 100-year rainfall 

 record at New Bedford, Mass., to determine the value of records of much shorter 

 duration are reported. 



It was found that a 5-year record based on the New Bedford record is too 

 short to give dependable results. The 10-year curve is more regular than the 

 5-year curve, but the 20-year averages vary from 4.2 per cent below the 100- 

 year average to 6 per cent above. " Hence, during the entire century it would 

 not have been possible to select any 20 successive years in which the average 

 rainfall would have deviated by more than 6 per cent from the 100-year average." 



Rainfall and production, L. McCooK {Agr. Qaz. N. S. Wales, 26 (1915), No. 

 5, pp. 3S9, 390, j)ls. 2). — Two charts, showing graphically the variations In 

 annual rainfall and yields of wheat and number of sheep in Queensland from 

 1873 to 1914, are given and briefly discussed. 



The charts show the depressing effect on wheat yield of the droughts of 1888 

 and 1902, especially the latter, which followed six years of low average rain- 

 fall. The drought of ISSS had little effect on sheep production, but that of 1902 

 was disastrous. There was slow but steady expansion of wheat production and 

 a more rapid increase in sheep production up to 1891. Thereafter wheat pro- 

 duction rapidly increased while sheep production declined. 



SOILS— FERTILIZERS. 



Soil experiments on the level prairies of northeast Missouri, M. F. Millee, 

 C. B. Hutchison, and R. R. Hudelson (Missouri Sta. Bui. 126 (1915), pp. 

 317-354, fiffs. 6). — These experiments were begun in one case in 1905 and in 

 two cases in 1907. The soil on which the experiments were made is known as 

 the Putnam silt loam, which includes practically all of the level prairie lands 

 of northeast Missouri. " The surface soil consists of a gray to dark gray silt 

 loam, from 8 to 10 in. deep, underlain by an ashy gray silt faintly mottled with 

 yellow. . . . Beneath this is a clay subsoil, beginning abruptly at an average 

 depth of about 18 in., which is made up of two layers. The upper layer is a 



