Gates, Tetraploid Mutants aud Chromosome Mechanism.-. |;;r 



chromosome-doubling and the various changes it entailed. This view 

 may be correct but it has certainly not yet been proved, and I 

 think it can he shown that most of the changes at least may have 

 resulted from a single original change, namely the doubling in 

 chromosome-number. De Vries cited the following characters as 

 unexplainable on the basis merely of the increased size of the nuclei 

 and cells: 1) The strong biennial habit. But this means merely a 

 slower rate of development under given conditions, and Keeble's 

 (1912) giant Primula also grew more slowly, a result which might 

 be expected to follow directly from the larger size of the cells, 

 and the slower rate of karyokinetic division. 2) The larger seeds. 

 It is difficult to see why this was cited, for the ovule and hence 

 the seed, is an organ which would obviously be larger if composed 

 of larger cells, as the yiyas ovule undoubtedly is. 3) The small 

 (i. e., short) fruits. The explanation here is not so obvious, but 

 the ovary at the time of fertilization is longer and stouter (see 

 Table II, p. 119) than in Laniarckicuia, again a direct result of the 

 larger cells; and in yiyas Sweden (see Table III, p. 126) the mature 

 capsule is also longer than in Lamarckiana, while it is shorter in 

 yiyas Italy and in lata. After careful comparative study, the explana- 

 tion of this is simple the length depends upon the amount of 

 sterility, or in other words upon the number of ovules which mature 

 seeds. If anyone examines a nearly mature capsule of yiyas or 

 lata they will find a comparatively small number of seeds and 

 a large number of undeveloped ovules. Part of these ovules doubt- 

 less fail to develope for lack of fertilization, but many of them (as 

 I have learned from cytological studies of oogenesis in lata] fail 

 to develope because the meiotic processes go awry. In lata the 

 sterility from this cause appears to be even greater than in yiyas. 

 The short length of the mature capsule in both these forms depends 

 upon the small number of seeds developed in them 29 ), and is there- 

 fore easily explained without recourse to another mutation al change. 

 The greater sterility of ovules, both in yiyas and lata, might be 

 expected to follow from the meiotic difficulties introduced by a) the 

 tetraploid, and b) an odd number of chromosomes. However, in 



29) There is a further interesting point, namely, that the seeds produced, 

 particularly in lata, are scattered through the length of the capsule, with many undeve- 

 loped ovules between them. The latter do not produce seeds because they are 

 incapable of being fertilized. On the other hand, when the fertilization of an ovary 

 in 0. rubrinervis and other forms is incomplete, through the failure of sufficient 

 pollen to reach the stigma, one finds almost invariably that the lowermost ovules 

 are the ones which are fertilized and develope seeds. Hence it appears that when 

 all the ovules are capable of being fertilized, the first pollen tubes must grow to 

 the bottom of the ovary, the next to the ovules next above, and so on to the top. 

 This behaviour cannot be explained by a summation of chemotactic influences from 

 all the ovules, for in that case the middle ovules would be the first to be fertilized. 



