726 



SCIENCE, 



[N. S. Vol. IX. No. 280. 



voting his time to what was formally an 

 ofScial work of the government of a foreign 

 country. 



This attempt to set forth in a few words 

 the scope and significance of twenty-five 

 years unremitting labor on the part of one 

 who would have made his mettle felt in any 

 sphere of activity he might have chosen to 

 enter may well appear to others, as it does 

 to us, extremely inadequate. The work of 

 Gill may fairly be called epoch-making in 

 a sense even better than that in which the 

 term is commonly used. If we find in it 

 no brilliant discovery to attract the atten- 

 tion of the public, it offers us what is yet 

 better ; improved instruments and methods 

 of research applied with such tireless in- 

 dustry, conscientious care, sound judgment, 

 and accurate knowledge of every related 

 subject as not only to expand our intellec- 

 tual horizon, but to supply the astronomer 

 of the coming generation a pattern which 

 he can study with profit to himself and ad- 

 vantage to his science. 



S. Newcomb. 



EXPERIENCE NOTES UPON PLOT EXPERI- 

 MENTS* 



Field experiments are not easy to plan 

 and very difiBcult in execution. Uncertainty 

 attends every step from the soil to the seed 

 — the cultivation to the harvesting of the 

 crop. 



During the past six years the writer has 

 had two acres under experimentation, and 

 no lesson has been more impressive than 

 that of the lack of uniformity of the soil. 

 An apparently even plot will vary in the 

 composition and texture of the soil, almost 

 from one foot to another. This may be due 

 to many causes, not the least important 

 being the rock strata, gravel beds, etc., that 

 underlie the soil. The surface is usually far 



•■■ Presidential address before the Society for the 

 Promotion of Agricultural Science, Columbus, Ohio, 

 August 19th. 



from level and the shaving down and filling 

 in to bring the surface to a grade only 

 emphasized the differences that already 

 existed, as results upon the experiment 

 grounds before mentioned abundantly 

 prove. 



The field itself should be laid off with 

 exactness, and this means the aid of the 

 surveyor or his instruments. The corner 

 stakes should be set deeply and never re- 

 moved, and those of each plot should be 

 established once for all. Unless this is so 

 the plots will move and the subdivisions 

 will vary, and shortly the whole area is 

 unsatisfactory, if not in confusion. 



Ideal experiment grounds should have 

 the same exposure. There are objec- 

 tions to perfectly fiat land, and there are 

 more to a rolling surface. The grounds 

 under the charge of the writer are upon a 

 slight incline, nearly uniform throughout 

 the two acres ; but even there a wash of 

 the soil is always troublesome during heavy 

 rains, and introduces an element of uncer- 

 tainty when soil treatment obtains upon 

 the plots. 



There should be no tree, bush or other 

 tall object upon or close to the trial grounds. 

 The shade of a single tree may do more to 

 disturb the course of an experiment with 

 sun-loving crops than the sapping of the 

 soil of plant food by the roots of the same 

 tree. 



This matter of the influence of shade has 

 been tested by means of lath by the 

 writer ; it is very great, and incidentally it 

 should play no part in the experiment 

 grounds. It is not safe to grow corn or 

 any like tall crop within many feet of any 

 small crops, as lettuce or spinach. The 

 scheme needs to be so planned that the 

 question of shade is reduced to a minimum. 

 This is one reason why the outside plants 

 are not comparable with these in the inte- 

 rior of the plot. 



A single back furrow or a dead furrow 



