8 MASS. EXPERIMENT STATION BULLETIN 195. 



Fig. 1 is a graphic representation of the variation indicated b}^ the 

 above figures. 



There was, unquestionabh', a great reduction in j'ield in 1915, 1916, 

 and 1917. Can we look to a specific soil trouble, pathogen, or method of 

 culture to account for this sudden and marked general decrease in yield? 

 It is hardly conceivable. A careful study of the yield in comparison with 

 the meteorological conditions, however, does furnish us wdth important 

 data, at least partially explanatory of the same. A studj^ of Plates I and 

 II (A and B), it is believed, will convince the most skeptical that the 

 weather conditions during anj^ given growing season determine to a great 

 extent the yield, and that the general reduction in 1915 and 1916, as well 

 as in other bad years, must be primarily attributed to these factors. 



In these plates the normal for this locality is represented as the straight 

 horizontal line or mean, and designated as (0). The variations of any 

 given period — in this case, monthly — above or bejow normal are 

 represented in black by the difference between the normal and that of any 

 given month. Variations above normal are represented above the line, 

 and below normal, below the line. Sunshine variation is given in hours, 

 temperature variation in degrees Fahrenheit (F°), relative humiditj^ in 

 percentage, rainfall in inches, and j'ield in pounds per acre. Each season 

 is divided into three parts, corresponding to the three principal months 

 during which maximum growth occurs, — June, July, and August. 



Studj'ing these tables it is found that there is a rather close correlation 

 between the various factors and the yield. Rainfall is, in Massachusetts, 

 probably the most important factor bearing on the yield, followed closely 

 by temperature and sunshine. The relative humidity apparently is not 

 of such importance, so far as the actual yield is concerned. Calling 

 attention to only a few of the more important differences will suffice. 



1915. — Sunshine slightly above normal in June, much below normal 

 in July and August; temperature, 1.5-2.0° below normal for the entire 

 growing season; relatii^e humidity above normal except in June; rainfall 

 practically normal in June, but from 4 to 5 inches in excess of normal 

 during July and August; yield, 480 pounds below normal! 



1916. — Here we have the reversal of conditions before mentioned: 

 sunshine much below normal in June and July, above normal in August; 

 temperature below normal for June, but above in July and August ; rain- 

 fall excessive except in August, when it was 2 inches less than normal; 

 yield better than in 1915, but still 180 pounds below normal. 



1917. — Here we have a partial return to more normal conditions, 

 but still somewhat abnormal; yield, oO-t- pounds below normal for stalk 

 tobacco. 



Going back still further it is possible to find analogous conditions in 

 other years. It is safe to conclude that if there is excessive rainfall during 

 the growing season, combined with low average temperature during the 

 first half of the growing season particularly, the yield per acre will be 

 small. 



