256 



NATURE 



[October 21, 1920 



.some on the male side appear to be inert and able 

 to carry only the recessive characters, and hence are 

 represented as RR, in contrast with the DR pair of 

 the female side. By this formula we can indicate the 

 behaviour of the several double-throwing strains. It 

 is, besides, becoming clear, I think, from recent 

 results that there is no "crossing-over" of these 

 factors on the male side in the F, cross-breds. But 

 the re^l difficulty is to explain why these factors are 

 confined to the female side in the ever-sporting 

 individual. This may result from aberrant behaviour 

 or from loss of chromosomes at some point in pollen 

 development. On this point I hope that evidence will 

 shortly be available. Failing such evidence, the pre- 

 sumption is that the elimination of XY (and in one 

 strain of VV) must have taken place frior to, and not 

 at, the moment of the maturation division. Morgan's 

 proposal to fit the pollen into his scheme for Droso- 

 phila by having recourse to hypothetical lethal factors 

 does not appeal to the observer, who finds the pollen 

 all uniformly good and every ovule set. Other 

 examples suggesting premeiotic segregation can be 

 quoted, notably cases among variegated plants and 

 plants showing bud sports, where somatic segrega- 

 tion appears to be of regular occurrence. 



Ovu! 



es 



Poll; 



It has been argued from time to time that any 

 scheme representing the mechanism of heredity 

 which leaves out of account the cytoplasm must 

 prove inadequate. This general statement has been 

 expressed in more definite form bv Loeb ("' The 

 Organism as a Whole," 1916), who hokls that the 

 egg cytoplasm is to be looked upon as determining 

 the broad outlines — in fact, as standing for the embryo 

 "in the rough," upon vi'hich are impressed in the 

 course of development the characteristics controlled 

 by the factors segregated in the chromosomes. The 

 arguments in favour of the view that the cytoplasm, 

 apart from its general functions in connection with 

 growth and nutrition, is the seat of a particular 

 hereditary process are mainly derived from observa- 

 tion upon embryonic characters in certain animals, 

 chiefly Echinoderms, where the inheritance appears 

 to be purely maternal. It has been shown, however, 

 that such female prepotency is no indication that in- 

 heritance of the determining factors takes place 

 through the cytoplasm. Other causes may lead to 

 this result. It has been observed, for exarnple, that 

 hybrid sea-urchin larvas, which at one season of the 

 year were maternal in type, at another wee all 

 paternal tn character, showing that the result was 



NO. 2660, VOL. 106] 



due to some effect of the environment. Again, where 

 the hybrid plutci showed purely maternal characters 

 it was discovered by Baltzer {Archiv jur Zelljorschung, 

 vol. v., 1910) that in the earliest mitoses of the cross- 

 fertilised eggs a certain number of chromosomes fail 

 to reach the poles, and are, consequently, left out of 

 the daughter nuclei. The chromosomes thus lost 

 probably represent those contributed by the male 

 gamete, for in both parents certain individual chromo- 

 somes can be identified owing to differences in shape 

 and size. .After this process of elimination, those 

 characteristic of the male parent could not be traced, 

 whereas the one pair distinctive of the female parent 

 was still recognisable. In the reciprocal cross where 

 the first mitosis follows a normal course the embryos 

 are intermediate in regard to the character of the 

 skeleton, thus affording proof of the influence of the 

 male parent. .Another type of case is found in the silk- 

 worm. Here a certain rate-character determining the 

 time of hatching out of the eggs has been shown to 

 e.xhibit normal Mendelian inheritance, the appearance 

 that it is transmissible by the female through the 

 cytoplasm alone being delusive. The eggs are always 

 laid in the spring. .According as they hatch out im. 

 mediately so that a second brood is obtained in the 

 year or do not hatch out for twelve months, the 

 female parent laying the eggs is described as bivoltin 

 or univoltin. Now the length of Interval before 

 hatching is obviously an egg-character, and therefore 

 maternal in origin. Consequently, when a cross is 

 made between a univoltin female and a bivoltin male 

 the eggs laid are not cross-bred in respect of this 

 character, any more than the seed formed as a result 

 of a cross is cross-bred in respect of its seed-coat, 

 which is a maternal structure. The silkworm mother 

 being univoltin, the eggs will not hatch out until the 

 following spring. The F, mother will, in turn, lay 

 eggs which again take twelve months to hatch, since 

 the long-period factor is the dominant. It is not 

 until the eggs of the F, generation are laid that we 

 see the expression of the character introduced bv the 

 bivoltin father. For some of the egg batches hatch 

 at once, others not for twelve months, showing that 

 of the Fj females some were univoltin and some 

 bivoltin, and hence that the egg-character in anv 

 generation depends upon both the maternal and the 

 paternal antecedents of the female producing the eggs. 

 Consequently, in the case of an egg-character the 

 effects of inheritance must be looked for in the genera- 

 tion succeeding that in which the somatic charac- 

 teristics of the zygote become revealed. We find, in 

 fact, that in almost all instances where the evidence 

 is suggestive of purely cytoplasmic inheritance, fuller 

 investigation has shown that the explanation is to be 

 found in one of the causes here indicated. The case 

 of some plants where it has been established that 

 reciorocal hybrids are dissimilar still, however, remains 

 to be cleared up. We know nothing as yet of the 

 cytology of these cases, and it is not improbable that 

 the interpretation may be found in some aberrant 

 behaviour of the chromosomes. .An instance in a 

 plant type where a definite connection appears trace- 

 able between chromosome behaviour and somatic 

 mipearance has been recently emphasised bv Gates 

 (New Phvfologist. vol. xix., iq2o), who attributes 

 the tieruliaritv of the lata mutation in Oenothera 

 (which has arisen as a modification at different times 

 from each of three distinct species) to an irrejjularity 

 in meiosis in the germ mother-cells whereby one 

 dauehter-cell receives an extra duolicate chromosome 

 which is lackins; in the sister-cell. The cell with 

 the extra chromosome fertilised by a normal germ 

 produces a lata individual. On the chromosome 

 view every normal fertilised egg contains a double 



