WHEAT IMPROVEMENT IN AUSTRALIA. 25 



Departmental action. 



Systematic attempts to improve wheats have been carried on by the South 

 Australian Department for many years past. The most complete scheme is 

 that introduced by Professor A. J. Perkins, Principal of Roseworthy Agri- 

 cultural College since 1892, and now Director of Agriculture, South 

 Australia. Professor Perkins bases his method on the assumption 

 that selection, to be effective, must be continuous and uninterrupted. The 

 broad principle underlying the method is that the best ears are selected from 

 the " selection plots," and are used for the following season's selection plot. 

 The process thus continues indefinitely and automatically, those ears 

 possessing in the highest degree the qualities sought for being alone selected 

 for further propagation in the selection plots. These plots have now been in 

 existence since 1904, so that, with some of the original wheats, the ninth 

 consecutive selection has been reached. Professor Perkins has very kindly 

 supplied the subjoined description, which will make his method readily 

 understood. 



" We start with a limited number of good ears selected from well-grown 

 plants in a standing crop, let us say King's White. The immediate progeny 

 of these ears (with the exception of the best picked from them in the 

 following season) I call King's White, first selection. The "best ears picked 

 from the best plants become King's White, second selection, and so on 

 indefinitely, the very best ears always becoming one selection in advance of 

 their immediate parents. At the present time we have reached our ninth 

 selection with some of the earliest wheats handled. 



" In any given year the. picked ears are sown in what I term selection 

 plots, i.e., in rows 24 inches apart, each grain being dibbled in one link 

 apart. Each ear is maintained separate under a number or letter, so that I 

 always have several strains of the same variety, many of which are discarded 

 in the course of time if they do not come up to expectations. In the same 

 way all our wheats are strictly pedigreed in that I can trace them all back, 

 strain by strain, to the individual original selections. 



" In picking out the best ears attention is paid to the usual points. First, 

 the plants must be thoroughly satisfactory, well grown, stooling normally, 

 find not lodged in any way. Secondly, the largest and best formed ears are 

 roughly picked in the field from the selection plots, and are subsequently 

 handled in the laboratory. Thirdly, as much as possible we give preference 

 to perfect ears, i.e., those carrying no empty shells. Fourthly, ears are 

 measured individually, and the number of spikelets counted; a record of this 

 is kept throughout. Only those most satisfactory in this direction are 

 retained. Fifthly, the ears are hand-threshed and the grain examined and 

 counted; the best again are retained. Sixthly, and finally, the grain of 

 individual ears is always hand-graded before sowing. 



"After picking out the best heads for next year's selection plots, we run a 

 stripper over the remaining plants, the grain from which goes into what I 

 liave called seed plots, extending over half an acre, 1 acre, or even at times 



