232 EXPERIMENT STATION RECORD. [Vol.38 



the rate of growth is considered to be fully as important under Connecticut 

 conditions us any increase in yield. 



" The highest yielding parents gave the highest yielding crosses, . . . but 

 . . . there was apparently no relation between the yield of the parents and 

 the increase in the yield of the cross. High average yielding parents gave as 

 large increases, when stated in percentages, as low yielding parents. There 

 was a tendency for the crosses whose parents differed in their ability to yield 

 to give the greatest increase. This is also shown by the fact that the dent 

 X flint crosses gave greater increases in growth than the flint X flint crosses. 



" These facts bear out the assumption that hybrid vigor is not the result of 

 an indefinite physiological stimulation, but merely the result of the bringing 

 together of [the] greatest number of favorable growth factors. Crosses between 

 \arieties of diverse type therefore possess a greater total number of favorable 

 growth factors than crosses between similar varieties, and hence give larger 

 increases when crossed." 



A statistical study of some indirect effects of certain selections in breed- 

 ing Indian corn, H. L. liiEiz and L. H. Smith (U. S. Dcpt. Ayr., Jour. Agr. 

 Research, 11 (1911), No. ^, pp. 105-146, flf/s. 24). — In connection with breeding 

 experiments with high- and low-protein and high- and low-oil strains of corn, 

 conducted at the Illinois Experiment Station and previously noted (E. S. R., 

 20, p. 531), the authors report observations upon what they term "the indirect 

 effects " of the selections by a statistical investigation of changes in certain 

 physical characters of the ears of corn, including length, circumference, weight, 

 and number of rows of kernels. The study involved the preparation of 476 

 distinct frequency distributions, and tabulated data are presented Showing 

 the frequency distributions with respect to the physical characters of the ears 

 for the four strains in the crop of 1914, and showing the type and variability of 

 the four strains with respect to each character studied for 11 crops, lOOH-lDla, 

 inclusive. Similar observations are recorded for the two-ear strains, 1908-1916, 

 inclusive, the erect- and declining-ear strains, 1907-191G, inclusive, and for the 

 high- and low-ear strains. 1907-191 G, inclusive, with considerable tabulated data 

 and numerous graphs illustrating the means and standard deviations for each 

 character studied in the different strains of corn. The results of the observa- 

 tions are summarized as follows : 



" It is found that four distinct types of corn as regards length, circumference, 

 weight of ears, and number of rows of kernels on ears are so well established 

 that we may assign orders of values to the means of these characters that per- 

 sist with but few exception in such changes of environment as have been ex- 

 perienced in 11 years of planting, from 1905 to 1915. While a few slight but 

 significant progressive changes have been noted, the selections for chendcal 

 composition from 1905 to 1915 have not changed decidedly the differences in 

 mean values of these characters. In fact, we are imable to assert with any 

 high degree of probability that the strains difler more or less with respect to 

 these characters during the second half of the period 1905 to 1915 than during 

 the first half. 



" The standard deviations of the strains do not differ nearly so nuich com- 

 pared to their probable errors as do the means, j;nd it is not in general nearly 

 so easy to discriminate among strains by the differences of standard deviations 

 as by the use of means. There is one marked exception to this, in that we 

 easily distinguish high-protein and high-oil from low-protein and low-oil strains 

 by the differences in the standard deviations in weight of ears. 



" No progressve change of consequence has taken place in standard deviations. 

 The coefficients of variability, in comparison to their probable errors, differ 



