WINNIPEG, 1909. 805 
at presetit uhknown. Both the Soft and the Dwarf wheats date back to 
equally remote times, both show a similar distribution in the past, and 
their general structural resemblance is sufficiently great to suggest a common 
origin. What it was we know not. Their primitive form will probably one 
day be traced back to a third species of wild wheat,and to an area not very 
far distant from that whence the Emmer and Hard wheats came. Here at 
any rate is ample scope for further exploration. 
The last of the wheats with which I have to deal is the Spelt proper, 
Triticum Spelta. Its origin has so far been quite obscure. It is mostly 
considered to have descended from Emmer, although the structure of its 
axis and its glumes lend little support to that theory, which probably had 
its origin in the early confusion of the popular names of the Spelt wheats. 
Spelt has so far never been found in any of the prehistoric settlements. 
Even its name Spelta was unknown until a.p. 301, when it appears for 
the first time in an edict by the Emperor Diocletian. It probably came to 
the Romans from the Germans, as Gradmann suggested in 1901. Since 
then, in a monograph, ‘ Der Dinkel (Spelt) und die Alamannen,’ he has 
produced strong evidence that Spelt was the staple grain of the Alamans, 
who brought it with them into South and West Germany from their old 
home east of the Elbe. From there it would have spread to the Alps, Italy, 
France, and Northern Spain. Gradmann’s-expression ‘east of the Elbe’ 
has evidently to be taken in a wide sense, as there is no grass in Eastern 
Germany or Western Russia from which Spelt could have descended. But 
still further east, or rather south-east, around the northern shores of tho 
Black Sea, whence the Alamans may very well have come originally, a 
species of Algilops occurs, A. cylindrica or Triticum cylindricwm, which 
comes structurally so near to the Spelt that I feel almost convinced that it 
is the primitive form of the latter. The usual slender forms of Triticum 
cylindricum are perhaps not so suggestive in this direction; but occasional 
stouter specimens approach very closely to it in the nature of the spindle, 
the texture and the cut of the outer glumes, and the curiously protruding 
inner or floral glumes. Moreover there is at Kew a specimen of Triticum 
cylindricum raised from seeds found among grain refuse in the Leith Docks, 
near Edinburgh, which might be taken for a pubescent but otherwise almost 
typical Spelt, except for the characteristic awns of the terminal spikelet. 
This theory of the descent of Spelt from Triticum cylindricwm does not 
exclude the possibility of early crossing with true wheats, as some ,of the 
peculiarities of Spelt suggest. For all we know the Leith casual may be 
a cross between T'riticum cylindricum and a pubescent true wheat, similar 
to the hybrid of Triticum Spelta and a pubescent unbearded wheat, described 
by Koernicke as Triticum NSpelta var. recens. 
To summarise briefly, we have traced the wheats to four primitive 
types: (1) the Kinkorn, to Triticum egilopioides, with its original home 
in Asia Minor and the north-eastern Balkans; (2) the Emmer and the 
Hard wheats, including the English and Polish, to Triticum dicoccoides, in 
Palestine; (3) the Soft, and probably also the Dwarf wheats, to a still 
unknown species, either in Syria or in Mesopotamia; and (4) the Spelt, 
to Triticum cylindricum, in an area extending from Bulgaria through 
Roumania to Southern Russia. From this standpoint the practical division 
into Spelt wheats and wheats proper breaks down entirely. Einkorn and 
Spelt proper, with their respective primitive forms, will have to stand as 
distinct species in any case; while it may be left an open question whether 
all the others should be treated under one or two species until we know the 
primitive form of the Soft and Dwarf wheats, and are able to gauge its 
taxonomic value as compared with Triticum dicoccoides. 
I have so far tried to remain on ground which allows us to work with 
tangible material and by way of comparing actual specimens. It will now 
