GRADES FOR COMMERCIAL CORN. S 
SECURING A REPRESENTATIVE SAMPLE FROM THE BULK. 
In the <ira cling of coniniercial corn no factor is of gre^^ter importance 
than the securmg of a sample representative of the bulk. Likewise, 
no factor is more frequently neglected. In the application of these 
grades to car-lot sliipments of corn it is recommended that not less 
than five probes with a suitable grain trier be taken in such a way 
that the composite sample thus secured will represent the average 
of the car as nearly as practicable. On cars not uniformly loaded, 
such additional probes should be made as, in the opinion of the 
sampler,- may be necessary to secure a representative sample. In 
cars that show distinct evidence of having been ''plugged," and in 
all cases of marked variation in the quality or condition of the corn 
in different parts of the lot being examined, a separate composite 
sample should be taken to represent each such portion. 
If only a part of the grain secured by the various probes is taken 
to a central ofhce for more careful examination and final grading, 
the mixing of the individual sample at the car should receive most 
careful consideration. Very satisfactory results can be secured by 
emptying the contents of the trier each time on a piece of canvas 
and, after all probes have been made, thoroughly mixing the samples 
on the canvas, finally rolling the sample on the canvas, somewhat as 
an expert would roll a cigarette, except that the canvas should be 
held by two opposite sides, which two sides should be securely fas- 
tened to a stick or rod. The larger composite sample can then be 
readily divided into two approximately equal parts by seizing the 
fold of the canvas from beneath with the thumb and index finger; 
then, empt3^ing one portion into the car, the other is retained for the 
office sample. 
Representative samples can not be secured by emptying the con- 
tents of the trier, after each probe^ on top of the grain, then roughly 
mixing and taking a portion thereof, usually including a part of the 
surface corn, as a composite sample for the basis of gracUng. Such 
samples not only fail to represent the bulk, but are misleading, 
especially from the standpoint of dirt and cracked corn. Likemse, 
composite samples made up by empt3"ing only a part of the contents 
of the trier into the can or sample bag can not, as a rule, be consid- 
ered representative. 
In the sampling of large lots of grain, such as occur in the loading 
of steamers, at least one representative sample made up of a series 
of samples from the various drafts should be taken for each 5,000 
bushels. 
In the sampling of ear corn, where the moisture content is the 
important factor, at least 20 representative ears should be taken at 
random for each 1,000 bushels. In wagon lots of 100 bushels or 
