HOWARD 9 



ill Avlilch they occur are regions of great aridity in which natural crossing is 

 common. It is possible therefore that they represent the result of such 

 crossing. As, however, at the present time, no dwarf wheats are found in 

 any of the localities mentioned, this would appear to be improbable. ^ 



In the conception of the variety, the definition of Koernicke^ has been 

 followed. A variety is considered to be a group of forms distinguishable 

 from other forms by some definite botanical or morphological character which 

 can be easily recognized in any well-grown indi\adual and converselv. the iji- 

 di^ddual forms comprised in the group should be indistinguishable among each 

 other by any such character. Botanical or morphological characters may be 

 defined as those which remain constant \vith change of environment or season 

 and which can be determined in the laboratory from properly developed 

 specimens. 



In contradistinction to botanical characters are agiicultuial or field, cha- 

 racters which cannot be distinguished in the laboratory or from individual 

 plants but can only be fully appreciated in the field in pure cultures, gro^Mi 

 side by side, under uniform conditions. Such chaiacters are time of maturitv, 

 height, susceptibility to rust. 



As the classification of the wheats in this paper is of necessitv based on 

 ripe specimens, the only characters which could be employed were botanical 

 or morphological ones and the classification therefore does not go bevond the 

 botanical variety. 



The characters employed by Koernicke in grouping his system of bota- 

 nical varieties were the following in the order given : — 



(a) Ears bearded or beardless. 



(6) Felted or smooth chaft'. 



(c) Colour of the chaif (red. white, or black). 



{d) Colour of the grain (red or white). 



An examination of a much larger number of forms than those which were 

 at the disposal of this observer and also the results of modern investigations 

 on the inheritance of characters in wheat have shown that the alternatives 

 enumerated above by no means cover the number of characters available for 

 the distinction of botanical varieties. To take the first case. " ears bearded 

 or beardless," it has been recently pointed out that the degree of bearding. 



' Since the above was written, wheats uf tlae type described have been obtained in the 

 F„ of a cross between a macaroni and an ordinary wheat. 

 2 Koernicke, 1. c. 



