No. 2, September, 1921] GENETICS 117 



lubricipeda, crossed with the yellowish-white type produces var. intermedia, which inbred 

 gives a 1 : 2 : 1 ratio (1 zatima, 2 intermedia, 1 lubricipeda) . A single factor, Z or z, differentiates 

 them, promoting the development of black pigment much more vigorously in the homozy- 

 gote than in the heterozj'gote. The great variability of intermedia and of zatima, and the 

 rapid -j- or — effect of selection in a series of either variety, leads to the conclusion that a 

 series of polymeric (modifying) factors influence Z in such wise that in Zz {intermedia) they 

 act strongly, in ZZ (zatima) less strongly, and in zz (lubricipeda) negligibly; the number of 

 these factors has not yet been determined. They control the distribution of the black pigment 

 in different ways and hence are not homomeric. Evidences of action of polymeric factors were 

 seen also in S. menthastri, Cerula furcula, and the Leucodonta bicoloria-albida-unicolora 

 series. — Consideration of polymeric factors in moths leads to the conclusion that species are 

 only apparently constant, consisting mainly of complex heterozygotes. Rare combinations of 

 genes give "varieties"; the rarest, "aberrations." Homozygous combinations of polymeric 

 factors, either dominant or recessive, producing the most extreme biotypes, occur in nature 

 only with extreme raritj^ — Critical examination of the supposed production of aberrations 

 in Arctia caja by exposing pupae to cold shows that this genetically highly variable form is 

 most unfavorable for such experiments, that the subsequent inheritance under normal temper- 

 atures by a few individuals of the aberrations supposed to be due to cold or heat is due to the 

 combination in these individuals of polymeric intensity factors and is not "inheritance of ac- 

 quired characters." — Selection for homozygous zatima soon reaches its limits; wing veins 

 remain white. Hence the completely black aberration, deschangei Depuiset, is probably 

 due to a special gene, not the result of cumulative selection of polymeric factors. — Zatima 

 was recorded in 17S2, intermedia not till 1890-1899, and the author assumes that the earliest 

 described zatima, unlike most melanics on first appearance, was homozygous [though dark 

 intermedia individuals are as black as zatima]. Polymeric factors so alter the phenotype as 

 literally to change white into black and lead to false conclusions regarding environmental 

 versus hereditary control of development, especially in man and other mammals which afford 

 relatively small numbers of individuals of the stock in question. — John H. Gerould. 



744. Frets, G. P. [Dutch rev. of: Bryn, H. Trondelagens Antropolcgi. (Anthropology 

 of Trondhjem.) K. Norske Videnskab. Selskab. 1917. Idem. Researches into anthropological 

 heredity. Hereditas 1 : 186-212. 1920 (see Bot. Absts. 7, Entry 1732).] Genetica3: G6-68. 

 1921. 



745. Harrison, J. W. Heslop. The genus Rosa, its hybridology and other genetical 

 problems. Trans. Nat. Hist. Soc. Northumberland, Durham, and Newcastle-upon-Tyne 

 5:244-298. PL 8-22,2 fig. 1921. — A study of a few closely related species of roses along with 

 their allied forms from various sections of the country shows that the types merge impercep- 

 tibly into each other, thus forming a "linked-up series of groups," with each group separable 

 from its neighbors by its own special characters. Further investigation shows that these 

 chains of groups are of common occurrence. Each chain is definitely divided from the next. 

 The writer considers each of these chains as being very close to a species in the Linnaean sense, 

 though he believes that most taxonomists would accord them the rank of a section or sub- 

 section. The members of the chain are here considered as Jordanian species, or "micro- 

 genes," as the writer calls them. The chains are called "section-species." Eight of these 

 section-species are recognized. Each species type is common to all section-species, seeming 

 to indicate an orthogenetic trend of development in the evolution of the rose. Thus a table 

 can be constructed, similar to Mendelejeff's Periodic Table, with the section-species as 

 vertical groups and the species types as horizontal groups. — A careful study of the pollen of 

 a number of roses reveals the fact that of 36 wild species and 2 garden hybrids produced from 

 wild species, only 4 possess perfect pollen, while 75 per cent of them have less than half their 

 pollen grains perfect, and about 40 per cent of them have less than 10 per cent of pollen capable 

 of fertilizing ovules. Seven forms produced practically no good pollen. The conclusion is 

 reached that pollen sterility is the outcome of latent hybridity. The pollen of R. rugosa and 

 R. cinnamomea, 2 fairly well isolated species, seems to be largely perfect. In default of out- 



