MICHIGAN ACADEMY OF SCIENCE. 209 



GRAUUS RABBIT-EAR ROGUES CROSSED WJTH NON-ROGUE- PRODUCING VARIETIES. 



From luy own observutioii.s as well as from the experience of other.s, it 

 would seem tliat not all peas with wide stipules give rise to Rabbit-Ear rogues. 

 From evidence that must be carefully verified, it would seem that Rabbit-Ear 

 rogues from Gradus, for example, when crossed with type plants of varieties 

 that do not produce Rabbit-Ear rogues, give results very different from those 

 obtained in crosses with type plants of Gradus. Gradus Rabbit-Ear rogues 

 wlien crossed with type itlants of a non-Rabbit-Ear producing variety seem to 

 give an Fi generation with stipules that are intermediate in width between 

 the rogue and tiie other parent. In the F:; generation a segregation taltes 

 place, so that in regard to shape of stipules, at least, there seems to be a wide 

 range of variation. Plants exceeding the wide-stipuled parent in stipule width 

 are found and plants about the same as the rogue, with regard to this char- 

 acter, are also produced. Between the two extremes lie a large number of 

 intermediate forms, many of wiiich repeat the splitting in the next generation.' 

 A few of the wide-stipuled segregates may breed true and some split into 

 various forms in the Fa. The Rabbit-Ear-like plants all seem to breed true. 

 No data have been obtained on the behavior of the various segregates when 

 crossed back with Gradus rogue and type plants. Crosses to verify the alxnc 

 statements were made in 1918. 



KESITLTS OF OTHER INVESTIG.VTORS. 



In 1915 Bateson and Pellew (4) reported the results of a .somewhat similar 

 study of rogues in Early Giant (E. G.), Duke of Albany (D. A.) and Ne Plus 

 Ultra (N. P. U.) Early Giant is a special strain of Gradus developed in 

 England by Sutton & Sons. The genetic behavior of the Rabbit-Ear type of 

 rogues was found to be like that of Rabbit-Ear rogues in Gradus. The authors 

 give additional facts concerning the hereditary behavior of various interme- 

 diate (intergrading) forms. Intermediate plants with type-like foliage and 

 rogue pods when crossed with type plants gave in the Fi generation progenies 

 either all rogues, intermediate and rogues, or type plants and intermediates. 

 In succeeding generations the various sorts of Fi plants behaved genetically 

 like similar plants which arose dc novo from type plants. 



It was found that the intermediates were of two classes, wJiich could be 

 .separated only by their genetic behavior. One class produced a larger uumiIxm- 

 of type plants than rogues and a second class was found in which the projior- 

 tions were reversed. In a later paper (5) it was noted that if the pods Iroiii 

 the upper and lower nodes of intermediate plants were sown .separately, the 

 greater number of rogues came from seeds produced on the upper part of the 

 parent plant. This was true in the case of intermediate forms which threw 

 a larger number of rogue than type plants in their progeny. Intermediates, 

 which prodnccil only n few rogues, with many type plants, showed no regu- 

 larity in lUc (lisiriiiuiion of rogues from sccd produced on didercut parts of 

 Ihti yluut. 



