588 Journal of Agricultural Research Voi. xix. No. h 



differently. If this regressive type were crossed on the intermedium form 

 the results, according to the hypothesis advanced, would not be the same 

 as when the Manchuria is so crossed. The Manchuria X intermedium gives 

 6-rowed, high-fertility, and intermedium segregates in a i to 2 to i ratio in 

 the Fj generation. The regressive 6-rowed X the intermedium- form 

 should give all classes, including 2-rowed forms, in the Fg generation. On 

 the other hand, if the regressive 6-rowed were crossed on the Svanhals, 

 no intermedium segregates would be expected. Unfortunately, neither 

 of these crosses has been made, but the latter condition has been met in 

 other crosses. Strains of Odessa and Reid Triumph when crossed on 

 Svanhals produced no intermediates, indicating that the regressive 6- 

 rowed AAbb does occur. The progeny of these crosses segregates in the 

 same way as the group which is heterozygous for the regressfive 6-rowed X 

 2-rowed in Table I. Incidentally, there is evidence that the strains used, 

 of both Reid Triumph and Odessa, originated as selections from a previous 

 hybridization. No evidence is at hand to indicate whether or not the 

 regressive 6-rowed type AAbb occurs elsewhere than in the progeny of 

 hybrids. 



Because fixed intermediates have been secured from hybrids in which 

 several old agricultural races of 6-rowed barley were used as one parent, 

 it is assumed that forms such as the Manchuria are the normal 6-rowed 

 forms. It is not known how frequently the regressive types occur among 

 our agricultural sorts, nor whether they are associated with other char- 

 acters than the failure to produce intermediates. Heritable variations 

 in the amount of fertility in the lateral florets of 6-rowed barleys have 

 been noticed, however, and since the conception of a regressive 6-rowed 

 type offers a possible explanation of their behavior, a statement con- 

 cerning these unusual varieties may be of interest. 



In 1909 the senior author noticed a decided tendency to sterility of 

 lateral florets in the agronomic varieties known as Mansfield and Summit. 

 In each case the earliest spikes to appear were normal 6-rowed spikes. 

 The lateral florets were highly fertile, long awned, and produced large 

 kernels. In the later spikes, especially where the plants produced several 

 culms, the lateral florets exhibited increasing sterility. Late spikes, in 

 which none of the lateral florets were fertile, were found in many plants. 

 One of these barleys, Mansfield, was the progeny of a cross between a 

 6-rowed and a 2-rowed barley, made at Guelph, Ont., in 1891. The 

 peculiar behavior of this variety was so striking that selections of fertile 

 and infertile spikes were made in 1909. These were planted in 1910, but, 

 as would be expected, there was no difiference in their inheritance. Owing 

 to seasonal conditions the percentage of spikes with no fertility in the 

 lateral florets was unusually large in 1910, the sterile spikes being 

 approximately equal to the fertile in number. Spikes from a single plant 

 are shown in Plate 105, D. It will be seen that the sterile florets are 



