34 



THE AGRICULTURAL NEWS 



February 3, 1912. 



True variation, us opposLil to Huctuation, is of two 

 kinds — continuous and discontinuous; the latter kind 

 is also known as mutation. Individuals exhibiting 

 continuous variation show an evenly graded series 

 lying between two extremes. Those which vary dis- 

 continuously may be collected into a few well-niarlied 

 groups. Individuals may show, in addition to variation, 

 small changes induced duiing the development of each 

 by the external influences to which it has been subject. 

 Such changes are known as fluctuations. A group of 

 individuals exhibiting some well-marked character, as 

 for example tallness, may show small differences, due 

 to fluctuation, within itself; but so long as the offspring 

 does not show changes in height that are quite outside 

 the limits of this fluctuation, that group is said to breed 

 true, and if the breeding can be continued for several 

 geneiations without the appearence of any marked 

 mutant forms, the line is said to be pure. On the other 

 hand, when ve,y marked changes in height occur in 

 the offspring of the first generation, some of thi- indi- 

 viduals of the group, if not all, must be regarded as 

 impure for the character of tallness. Mendel s princip- 

 les apply only to very simple characters such as tallness 

 or dwarfness, seed characters, flower colour and the 

 like, when these vary discontinuously; they show what 

 off'spring may be expected to arise when individuals, 

 breeding true to one character, are crossed with those 

 breeding true to its opposite, and when the residting 

 seedlings are fertilized among themselves. 



The principles also show how the characters are 

 distributed, when an individual breeding true to the 

 presence of one desirable attribute and the absence of 

 another, is crossed with a second individual showing the 

 absence of the first and the presence of the second. It 

 is from such a cross that new forms, breeding true to 

 both desirable characters, may be synthesized. It may 

 be stated finally, that continuous variation is believed 

 to be caused b}'^ large number of indeterminable fac- 

 tors, while mutation is probably due to one or a few defi- 

 nite determinants,and that the inheritance olcontinuous 

 variation is different from that of discontinuous, and 

 does not follow iMendel's laws; while fluctuation is not 

 inherited at all. 



In order to commence Mendelian experiments on 

 a Sound basis, it is necessary to have a knowledge of 

 the nature of the variation in the characters chosen, 

 a certainty of the purity of the parent strains as re- 

 gards those characters, and a knowledge of the simple 

 nature of the characters themselves, and of the probable 

 number of determinants which go to form each; in shorty 

 to p<>.ssess a satisfactory analytical basis from which to 

 commence synthesis. 



In the production of seedling sugar-canes in the 

 past, it has not been necessary to attach any great im- 

 portance to an exact knowledge of parentage. In the 

 earliest work, the parents were sometimes quite un- 

 known. Later, the female parent was known, but the 

 male was uncertain; while only recently have canes 

 been bred, for sevei'al generations, whose parent- 

 age is known on both sides throughout. Naturally, 

 also, most seedling canes must be regarded as complex 

 hybrids until definite information is avnilable to ci;)n- 

 tradict this assumption. Under such circumstances, 

 there is naturally no very precise understanding of the 

 type of variation exhibited by the different characters 

 of the sugar-cane; furthermore, no strain of canes could 

 be selected which is known to breed true to any char- 

 acter desireii. Thus the former records cannot supply 

 two of the items necessar\' for the analysis postulated^ 

 and at ^Jresent there is );o other source from which the 

 knowledge cbuld possibly be cbtained. 



As regards the characters themselves, those most 

 carefully studied and recorded are of purely economic 

 importance, and a very short consideration will show 

 that they are either too ill-defined or too complex in 

 themselves to permit of the expectation that they will 

 behave as Mendeliaa units. If, for example, the sucrose 

 content be considered, it apjJcars possible that it is sub- 

 ject to continuous variation, though, on the other hand, 

 certain canes exhibit a fairly constant value for this 

 quality. It is not, moreover, sufficiently defined 

 to enable a judgment to be formed as to the actual 

 nature of its variation, since errors arise in the inter- 

 pretation of the chemical analysis upon which it is 

 based, owing to differences in the ripeness of canes, to 

 the extent to which they have dried before being 

 crushed, and to other external causes. Possibly, if the 

 character were expressed as concentration of sugar in 

 the sap of the storage cells of any strain when quite 

 ripe, it might be found to be definite, and subject to 

 discontinuous variation. At present, such a character 

 is not likely to serveas a Mendelian unit. Again, ton- 

 nage of cane per acre may be an effect of the production 

 of many shoi-t canes to the stool, or of a few long shoots, 

 as w-ell as of intermediate conditions: so that any mathe- 

 matical valuation of this is not in reality a definite 

 character at all. Disease resistance is another vague 

 character, not properly defined or understood, that may 

 be inversely rlependent on sucrose content of juice. 

 The employment of other economic characters is sub- 

 ject to similar objections. Thus it would appear that, 

 in every way, the necessary analytical data are not 

 forthcoming. 



