2 BULLETIN 1386, U. S. DEPARTMENT OF AGRICULTURE 
the sorgos are used for forage, the requirements are evidently not sd 
exacting, more latitude being permissible in the proportions of the vari- 
ous constituents contributing to the feeding value. Sugar is not pro- 
duced commercially from this plant at the present time, but the 
sugars — sucrose, dextrose, and levulose — are the essential components 
of sirup made from its juice, and certain other components also con- 
tribute to flavor. In the United States at various times varieties have 
been examined from a chemical standpoint, and a number have also 
been tested experimentally and by growers regarding the actual yield 
and quality of sirup that may be obtained from them by the usual proc- 
ess of concentrating the juice, so that it is well known that there are 
marked differences in respect to sirup-making properties between some 
of them. .V number of examples of varieties between which differ- 
ences in sugar content are considerable, though visual differences are 
small or at least frequently not recognized by growers, might be 
cited. The following example is pertinent, as it shows that exten- 
sively cultivated kinds which differ widely as to saccharine proper- 
ties are sometimes similar. Two awnless varieties, both of which 
have semicompact panicles and black or nearly black glumes, have 
shown in recent tests an average difference of 3 per cent in the 
amount of soluble solids and something over 3 per cent in the quan- 
tity of sugars in their juices. This is equal to an increase of more 
than one-sixth in solids and over one-fourth in sugars. There are 
many others, differing in visual characters that are not usually recog- 
nized by growers and seedsmen, which are so diverse in sugar con- 
tent and other economic characters that it would be much more 
profitable to cultivate only certain ones and to discard all the others. 
Xew sorts which differ more or less in appearance are continu- 
ally being encountered. Some of them are grown only in particular 
localities; others are more widely cultivated and may be often 
found in communities in various States. Practically all seedsmen 
handle the ^cad of one or more varieties, and they become dissemi- 
nated in this way. Some of them present the appearance of much 
homogeneity, but in some cases these are in reality mixtures of a 
number of somewhat similar forms. Others are very evidently im- 
pure. It is usually possible by continuous roguing to eliminate 
the more conspicuously diverse types from these commercial varie- 
ties. About 20 of the varieties upon which the present study is 
based were inbred for a series of years, and it was thus possible 
largely to eliminate forms which did not conform to the prevailing 
type. Although such a practice might not always be feasible com- 
mercially, as growers would doubtless find it somewhat difficult to 
make fine distinctions and might find it especially difficult to main- 
tain pure races once they were obtained, the writer knows of one 
commercial firm which practices breeding based on tests of the 
sucrose content of individual stalks, the heads from the best-testing 
stalks being saved and the seed planted for increase. It would thus 
seem that with some knowledge of varietal characters it would be 
practicable to isolate and cultivate comparatively pure races which 
show superiority in sugar content or other economic characteristics. 
As varietal characters become better known, it should become increas- 
ingly feasible for both growers and seedsmen to distinguish the 
varieties grown and for the growers to make finer distinctions be- 
