312 



NA TURE 



[March 9, 1922 



the growth exhibited and the nature of the differentia- 

 tions displayed by the tumour cells. The discovery 

 of the transplantability of tumours of the lower animals 

 has provided much material for this line of research, 

 but the many attempts made to fix on any one out- 

 standing character of tumour cells differentiating them 

 sharply from normal cells have been unsuccessful. As 

 before, we are confronted with the unexplained and 

 unco-ordinated powers of proliferation shown by the 

 tumour cells. The discovery that animals could be 

 rendered resistant to transplanted tumours raised hopes 

 that it might be possible to elicit an immunity towards 

 cancer in an animal affected spontaneously, but these 

 hopes are now considerably abated. 

 A start has also been made to ascertain the food 



requirements, general and special, of the tumour cells, 

 but these experiments are still too slightly advanced 

 for us to know whether any result of positive value 

 will be obtained. 



Research into the treatment of cancer other than 

 surgical has produced many empirical experiments and 

 observations, but, apart from the extended knowledge 

 of radio-therapy, nothing of importance has come to 

 light. In the field of radio-therapy, the manner of 

 action of the rays used, and the way in which they induce 

 destruction of cancerous cells, still offers an unsolved 

 problem of high importance. In conclusion, it may be 

 predicted that progress in cancer research will in 

 large measure be closely co-ordinated with that in the 

 ancillary sciences. 



The Mechanism of Heredity.^ 



By Prof. T. H. Morgan, Columbia University, New York City, U.S.A. 



III. 



Further Relations between Chromosomes and 

 Heredity. 



IN examining the chromosomes for a stage when 

 " crossing-over " might be possible, we turn 

 naturally to the time when the members of each pair 

 come together. This occurs once in the history of 

 every germ-cell. In many accounts it has been shown 

 that the members of each pair come to lie side by side 

 throughout their length. Even more interesting is 

 the fact that just prior to this union the chromosomes 

 have spun out into long, thin threads. There are also 



several detailed accounts showing that at this time 

 the two chromosomes of each pair may actually twist 

 about each other in one or more turns (Fig. 16). They 

 then come to lie side by side and appear as a single 

 thread that shortens preparatory to entering upon the 

 first maturation division. Here, apparently, we find 

 realised a condition that might make interchange 

 possible between the members of a pair of chromosomes, 

 for if the threads fuse where they cross each other and 

 the ends on the same side unite, the interchange of 

 pieces will be accomplished. From the nature of the 



• Continued from p. 278. 



case it would be almost impossible to demonstrate 

 that the twisted threads do break and make new 

 unions at the crossing point. It is true that there are 

 certain later stages that lend, perhaps, some support 

 to the view that breaking and reunion have occurred, 

 as Janssens has pointed out, but it cannot be claimed 

 that this evidence does more than give, on such an 

 assumption, an account consistent with certain con- 

 figurations he describes. Here the case must rest for 

 the present. The genetic evidence is clear and far in 

 advance of what the cytologist is able to supply. But, 

 nevertheless, it is very important to find that, so far 

 as the cytological evidence goes, it furnishes a great 

 many of the facts essential to the kind of process that 

 the genetic evidence calls for. 



The Number of the Linkage Groups and the 

 Number of the Chromosomes. 



When Sutton in 1902 directed attention to the fact 

 that in the behaviour of the chromosomes at matura- 

 tion there was supplied a mechanism for Mendel's two 

 laws, it was evident that the number of independently 

 assorting hereditary characters would be limited to 

 the number of the chromosome pairs characteristic 

 of each species of animal and plant, provided the 

 chromosomes remain intact from generation to genera- 

 tion. The integrity of the chromosome was held, in 

 fact, by a few leading cytologists at that time, notably 

 by Boveri, on evidence which, if not complete, was 

 the best then obtainable. In the circumstances, the 

 later discovery of the agreement between the number 

 of chromosome pairs of Drosophila melanogaster and 

 the number of its linkage groups was of paramount 

 importance for the chromosome theory. In this 

 species the number of known hereditary characters is 

 so large (more than 300 in all) that this relation can 

 scarcely be due to a coincidence, especially when the 

 whole evidence concerning chromosomes and heredity 

 is taken into account. 



It is true, with the possible exception of the garden 

 pea (where there appear to be as many independently 



NO. 2732, VOL. 



109] 



