86 



NATURE 



[September 15, 192 1 



deficiency or excess of these substances may pro- 

 duce abnormal results. The discovery of these 

 hormones, as they have been termed by Prof. 

 Starling, must have a profound influence on our 

 ideas as to the mechanism of heredity. Their 

 significance from this point of view was, I believe, 

 first pointed out by Mr. J. T. Cunningham many 

 years ago. It seems at least possible that chem- 

 ical substances of a like nature may exist in the 

 germ-cells and exercise a profound influence upon 

 their development. 



With this possibility in view, let us again 

 examine the egg-cell at the very commencement 

 of its . development into a multicellular body, at 

 the moment of its division into the first two 

 daughter-cells, and let us concentrate our atten- 

 tion upon the nucleus, which always divides first. 

 As it prepares itself for this important event a 

 number of peculiar bodies called the chromosomes 

 make their appearance, apparently by concentra- 

 tion of previously scattered granules of chromatin 

 substance, so-called because of the way in which it 

 can be stained by certain dyes. At the same time, 

 a spindle-shaped arrangement of threads becomes 

 manifest and the nuclear membrane disappears, 

 so that there is no longer any sharp division 

 between nucleus and cell-body. The chromo- 

 somes, often varying in shape and size amongst 

 themselves, but definite and constant for each 

 kind of organism, arrange themselves across the 

 middle of the spindle. Then each splits into two, 

 and one half moves away from the other 

 and towards the corresponding end of the 

 spindle. We have now two groups of daughter- 

 chromosomes, and around each group a new 

 nucleus is constituted. Then the protoplasm 

 of the cell-body divides into two parts, and two 

 complete cells are formed, each with its own 

 nucleus. 



The process is really far more complicated than 

 this brief and inadequate description might lead 

 you to suppose, but the essential feature seems 

 always to be the behaviour of the chromosomes. 

 . It is very evident that the protoplasm of which 

 these are composed must be of the utmost import- 

 ance to the organism, and that it is necessary 

 that it should be very accurately divided between 

 the daughter-cells every time cell-division takes 

 place. This phenomenon of mitosis, as it is 

 termed, is of almost universal occurrence through- 

 out the animal and vegetable kingdoms, not only 

 in the early divisions of the egg-cell, but through- 

 out the entire life of the organism, whenever cell- 

 division takes place. It is clearly a contrivance 

 bv which a certain material substance — a par- 

 ticular kind of living protoplasm — is accuratelv 

 distributed amongst the progeny of a dividing^ cell 

 In other words, it is part of the mechanism of 

 inheritance. 



Let us now turn aside for a moment and glance 

 very briefly at another and totally different line of 

 evidence, leading to results which confirm and 

 explain in a very remarkable manner those which 

 we have already arrived at. I refer, of course, 



NO. 2707, VOL. 108] 



to the modern experiments in the breeding of 

 plants and animals, undertaken under the influence 

 of what is frequently termed the Mendelian school. 

 It is utterly impossible to do justice to these 

 wonderful experiments in the time at our disposal. 

 1 would point out, in the first place, however, that 

 they have led quite independently to the striking 

 conclusion that there must exist in the protoplasm 

 of the germ-cells definite material entities — the 

 so-called Mendelian factors — which are in some 

 way or other responsible for the appearance in the 

 adult organism of special features — the so-called 

 unit characters — capable of being handed on from 

 one generation to another by the process of 

 heredity. Assuming them to be located in the 

 chromosomes, the behaviour of these factors in 

 inheritance, the permutations and combinations of 

 unit characters which arise in cross-breeding, can 

 be adequately explained by the behaviour of the 

 chromosomes actually observed at certain critical 

 periods of the life-cycle. 



Take, for example, the colour of the human 

 eye. If a certain factor, or combination of 

 factors, alone be present in the germ-plasm, the 

 eye will be blue or grey, but the addition of 

 another factor may cause it to be brown, and the 

 average results, as regards eye-colour, of mating 

 pure blue-eyed and pure brown-eyed individuals 

 can be confidently predicted. The occasional ap- 

 pearance of an extra thumb or finger upon the 

 hand, which is well known to be a heritable char- 

 acter, transmitted with great regularity from 

 parent to child, is again supposed to be due to 

 the occurrence of a corresponding factor in the 

 germ-plasm, and so on with a whole host of char- 

 acters that have been carefully investigated by 

 means of breeding experiments in recent years. It 

 is important to note that these characters seem to 

 bear no purposeful relation whatever to the well- 

 being of the organism in which they occur. They 

 are often extremely insignificant, and a large pro- 

 portion of them must undoubtedly be regarded as 

 abnormalities. It is a mere matter of chance 

 whether they happen to be useful, neutral, or in- 

 jurious. 



We cannot attempt to discuss, or even state, 

 however briefly, the evidence upon which this 

 factorial hypothesis rests. Suffice it to say that 

 it seems to afford the only possible explanation of 

 the results of the long series of experiments in- 

 augurated in Austria by the classical work of 

 Mendel in the middle of the last century, carried 

 on by Bateson, Punnett, Biffen, and others in 

 England, and culminating in the brilliant in- 

 vestigations of the American school of geneticists 

 under the leadership of Morgan. 



The investigations of Prof. Morgan and his col- 

 leagues have gone so far as to demonstrate con- 

 clusively, albeit indirectly, not only that the Men- 

 delian factors must be located in the chromosomes 

 of the nucleus, but also that they must be arranged 

 in each chromosome in a perfectly definite manner. 

 These observers have even prepared maps of 

 chromosomes showing the arrangement of the 



