November 18, 1904.] 



SCIENCE. 



681 



the maximum amount the soil should contain, 

 except for short intervals, in order to secure 

 the best growth. 



!But even if the author's conclusions be not 

 correct regarding the cause of the flow of 

 water in the experiment, the line of investi- 

 gation is important in that it has provided a 

 means of securing water from field soil, per- 

 haps, in a somewhat more concentrated con- 

 dition, than occurs in natural drainage and 

 permits the sample to be taken where its 

 history may be very definitely known; and it 

 is to be hoped that they and others will apply 

 the method in investigating the character of 

 soil extracts thus obtained. We regard it ex- 

 tremely doubtful, however, that either the 

 concentration or the composition of solutions 

 so procured will be found to be the same as 

 that which closely invests the soil grains or 

 the root hairs at the same place and time. 

 Certainly, if the movement is a capillary one, 

 the observations recorded in the bulletin* on 

 the ' Movement of Water-soluble Salts in 

 Soils ' indicate that very notable changes in 

 the composition of the solution may take 

 place as a result of the translocation. Our 

 own observations also show that when only 

 small quantities of some solutions are forced 

 through such filters the concentration may be 

 measurably changed. F. H. King. 



Madison, Wis., 

 November 2, 1904. 



HYBRID WHEATS. 



To THE Editor of Science: In my original 

 paper on hybrid wheats (Bull. 115, O. E. S.) 

 the second generation of crosses between long- 

 head varieties {Triticum vulgare) and club 

 wheats {T. compactum) were divided into 

 long, short and intermediate heads, these three 

 types occurring in the proportion 1:2:1. 

 Subsequent examination of later generations 

 of these wheats, all of which continue to obey 

 Mendel's law quite strictly, leads me to believe 

 that the short head of the club wheats is 

 really a dominant character, and that the 

 apparent intermediate character of the heads 

 of the heterozygotes of the several generations 



* ' Investigations in Soil Management,' by the 

 author, Madison, Wis. 



of hybrids is due to the greater vigor of the 

 heterozygote individuals. It was found in the 

 third and later generations that the long and 

 intermediate heads could be separated with- 

 out error, as shown by the purity of the long 

 heads next year. But there were many small 

 errors in separating the intermediate and 

 short heads. If the latter separation had been 

 perfect, the short-head type should have re- 

 produced true to type. But in a majority of 

 the plats supposed to contain only short heads, 

 a few long and about twice as many inter- 

 mediate heads were found, indicating that in 

 most cases one or more intermediates had 

 been selected with the shorts the previous year. 

 When the seed of each plant was kept sepa- 

 rate, this difficulty disappeared, each plant 

 behaving either as a pure short head type or 

 as an ordinary heterozygote. 



One of the most interesting results which 

 subsequent study of these hybrids has brought 

 to light is the apparent effect of hybridization 

 on the variation of single characters. For 

 instance, the length of head of the long-head 

 parent of the hybrid is fairly uniform, but in 

 the hybrid this character varies between wide 

 limits. The same is true of the length of the 

 short or club heads. In the parent club va- 

 riety the heads are fairly uniform in length; 

 but in the pure short-head progeny of the 

 second and later generations some of the 

 heads were less than one fourth the length of 

 the ordinary club heads. This induced varia- 

 bility of a character which has recently passed 

 through the stage of what we may call hetero- 

 zygosis probably accounts for the errors made 

 in separating the intermediate and short heads 

 above referred to. It is also of capital im- 

 portance to the practical breeder. Those who 

 are so situated as to attempt it will find an 

 interesting problem in the effect of selection 

 in fixing these variable characters, should the 

 power of yielding large quantities of seed be 

 rendered highly variable by hybridization, and 

 should we be able to fix unusual yielding 

 power thus induced, we could establish races 

 of great economic value. 



W. J. Spillman. 



U. S. Department of Agriculture. 



