Seed Wheat. 



23 



of this variability of conditions one may cite the difference in the supply of natural water 

 due to various causes. The rain may not fall evenly. Eminent meteorologists assure us that 

 rainfall is very capricious. The rainfall at places only a few feet apart may be quite 

 appreciably different, and under certain conditions of dryness this variation in the fall 

 probably exerts a considerable influence in causing variations in plant growth. Apart 

 from this, it needs no great experience to cause one to question, even when the rain falls 

 evenly, whether it becomes evenly distributed in the soil. Suppose the case of a short 

 shower, in which the precipitation is sharp for a few minutes ; such showers as are not at 

 all uncommon. How is it possible to know that the unavoidable variations in the surface 

 of the soil and variations in the underground condition of the soil do not cause an uneven 

 distribution of this water to the plants of a crop ? Let anyone watch the surface 

 distribution of water during a short shower, as the author has often taken the trouble to 

 do, on ordinary cultivated ground, and he will soon see, if the rainfall becomes sharp, how 

 unevenly the small surface currents distribute the water that afterwards soaks more or 

 less completely into the ground. Thought and observation of this kind have led me to 

 be rather cautious about making comparisons among plants and crops grown under field 

 conditions. 



It seems to me that the history of crop experiment work shows a multitude of 

 instances where caution of this kind has remained profoundly unthought of, or else has 

 been thrown to the winds, and I cannot help thinking that much of the admittedly 

 unsatisfactory nature of field experiment work reported is due to this lack of thoughtful 

 caution, combined with a readiness to report the results of a single year's experiments 

 in terms that do not point clearly to the low value that often attaches to an isolated 

 field experiment. 



The late Sir John Laws once said to the author, in a conversation on the underlying 

 principles of agricultural experiment work, that he early formed the habit of thinking 

 long over an agricultural theory or experiment before expressing an opinion. One sees 

 in that remark a grain of the wisdom that lias made Rothamstead so great a landmark 

 in the history of agriculture. No doubt it was this " long thought " that led to the 

 absence of rash expressions of opinion connected with Rothamstead work. 



Most of us might take a leaf from Sir John's book in this respect. To think long and 

 to say little, to repeat experiments again and again so as to be able to compare and verify 

 over and over, and yet say little ; to make endless excursions among the logical pitfalls 

 that are the special danger of the interpreter of field experiments, and yet to say little ; 

 or, as Sir John Laws put it, " to think long" before expressing a definite opinion on an 

 agricultural experiment or theory, this must certainly be the best possible course to adopt. 



Reasoning thus I have been led to abstain from calling attention to any comparisons 

 other than those it is possible to make between the adjacent rows of the seed-wheat 

 experiments now to be described. 



Seed Wheat: Large Plump versus Small Shrivelled Seed. 



The first of the systematic trials to ascertain the relative seed value of large plump 

 grains and small and shrivelled grains were undertaken in 1894. The results of these 

 trials are given in the following table : 



FIRST TRIAL LARGE PLUMP VERSUS SMALL SHRIVELLED SEED. 



