Referate. 291 



occurrence of a 3 : 1 ratio when a 1 : 1 would be expected, and the appear- 

 ance of dominants in the offspring of recessives. To explain these peculiarities 

 the hypothesis is offered that the plants possess a double set of factors along 

 with the double series of chromosomes. Further experiments are required 

 to confirm this inherently probable hypothesis. It may be pointed out that 

 the cases of duplicate factors for a single character, such as the red glumes 

 in the wheat of Nilsson-Ehle, may be due to a similar mutation or ger- 

 minal change having occurred independently in two or more chromosomes 

 belonging to different pairs. Gates. 



IIhs. H. The origin of < Capsclla bursa-p.astoris arnclinoidcii. 



Amer. Nat. 4S: 193-235, Apr. 1914. 



In cultures of ('(ipnella bursa -pcistoris derived from plants growing spon- 

 taneously in tlie greenhouse, the author was able to distinguish the reviewer's 

 two forms, C. hi), rhomboidca and C. bp. nimplex, as well as several types not 

 previously described. The most striking of the latter forms was one with 

 enlarged cotyledons, very narrow, unlobed, strap-shaped leaves, short flower- 

 stems, and almost completely sterile flowers. This form he names C. bp. 

 arachnoidal. Because of its sterility the author was unable to test directly 

 the genotypic constitution of this striking new type, but by well directed 

 experiments with collateral forms he was able to produce the same tyjie again, 

 and to show that it is homozygous with respect to a gene A'^ which decreases 

 the relative widtli of the leaves. When X is absent, as in rhomboidca and 

 ssimph'x, the earlier leaves are about twice as long as broad: when ^V is pre- 

 sent once, i. e. heterozygous, the early leaves are 2'5 to 3 times as long as 

 broad, and when homozygous the leaves are strap -shaped. Only in the latter 

 case is the fertility also affected. The gene N may also be defined as a 

 counteractor of the rlmmhuidca- gene, B. When AMs heterozygous the action 

 of B is only [lartially inhibited, but the inhibition is complete when iVis homo- 

 zygous. The author was usually able to distinguish the heterozygous forms. 

 He describes these in detail, assigning to them trinomials as follows: 



hbNn =: C. bp. attenuala 



BbNn = C. bp. SdchdUana 



BBXn=z C. bp. Trrlea.icana 

 The author also finds heterozygous rhotnboidea (Bbnn) frequently distin- 

 guishable from the corresponding homozygous form, as the reviewer has also 

 found, but to this heterozygote he gives no name. The desirability of giving 

 trinomial names to heterozygous biotypes may well be questioned, especially 

 in material so easily modifiable by slight variations in the environment. 

 The author confesses that he is frequently unable to classify these forms 

 except by a study of their offspring. The genotypic formulae seem to be 

 the most satisfactory designation for such forms. The results of the genetic 

 experiments leave no doubt of the essential correctness of the author's 

 analysis, but the manner in which a general deficiency in the number of 

 arachiwiilea individuals is explained, is a confession of an avoidable weakness 

 in technique that the author should improve; unintentional selections from 

 the seed-pan may be effectively guarded against if appropriate methods are 

 adopted to that end. The author points out that covering the inflorescence 

 is unnecessary in genetic experiments with Capsella, but this process is so 

 easily carried out with this species that one is not justified in taking the 

 risk of an occasional unintentional cross. The reviewer has frequently found 



