'252 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



(4) To determine the best cultural methods. 



The experiments were chiefly concerned with grains, wheat, oats, 

 and barley (winter and spring varieties of each), besides which 

 promising results were obtained with alfalfa, while broom-grass, 

 sorghums, peas, corn, potatos, and sugar-beets were also tested. One 

 of the grains tested was Emmer " (Black winter var.), little known 

 in the United States, botanically closely related to wheat, but resembling 

 barley, and valuable chiefly as a stock food. It resists draught and has 

 proved extremely hardy in Utah. It produces a high yield and is 

 much relished by horses, hogs, and sheep. — C. H. L. 



EelwOPms (Gard. Mag., No. 2954, June 11, 1910, pp. 455).— 

 Experiments on various methods of destroying eelworms have been 

 carried on by Dr. G. E. Stone, of the Massachusetts experimental 

 station. The use of lime and formalin were found to have little result, 

 and the latter is very injurious to plants. 



More success was obtained by sterilization with steam pipes to at 

 least 180°. Freezing the soil for a length of time destroys them, also 

 desiccation, but this is difficult to apply if there is any depth of soil. 



In Germany a catch crop like mustard or rape has been used, 

 the' plants being pulled up and destroyed. Flooding of soil or manure 

 to be used destroys many. — E. B. 



Experimental Results, Interpretation of. By J. B. Wood 



and F. J. M. Stratton (Jour. Agr. Sci. iii. pt. 4, pp. 417-440).— After 

 drawing attention to the need for caution in interpreting experimental 

 results the legitimacy of averaging results is discussed. The method 

 of calculating the probable error is described and its meaning 

 explained. 



It is found that the probable error in field experiments is about 

 5 per cent, of the crop. This figure is independent of the size of the 

 plot provided it is l-80th acre or larger. It is useless trying to 

 measure differences less than about 20 per cent, by comparing single 

 plots whatever their size. If it is desired to measure small differences 

 the number of plots must be increased either by duplication several 

 times in the same experiment, or by repetition of the experiment at 

 several stations, or for several seasons. If the precision desired in 

 percentage difference between yields is 20 per cent., then 1 plot may 

 be compared with another; if 15 per cent., then 2 plots are rejquired; 

 if 10 per cent., then 4 plots ; if 8 per cent., then 6 plots ; if 6 per cent., 

 then 10 plots ; if 4 per cent. , then 23 plots ; if 2 per cent: , then 91 

 plots. 



The whole paper is of great importance and very suggestive to 

 experimenters in all fields. — F. J. C. 



Felicia petiolata (Aster petiolatus). By D. Bois {Rev. Hort. 

 August 1, 1910, pp. 550-2; one illustration). — From the illustration 

 and description this plant would appear to be well worthy of attention 

 for sunny rockeries or other situations where a moderately rampant and 



