WORK OF THE CHEMICAL LABORATORY, 
By N. W. Lord. 
INTRODUCTION. 
In (his preliminary report only the mine .samples and the labora- 
tory car samples of the coals are considered. 
Numerous analyses made in connection with steam tests, producer- 
gas tests, coking tests, etc., representing the portions of coal used 
in these tests, are published in the several reports giving the results of 
the tests and are not repeated here. 
SAMPLING. 
The mine samples were taken by the agents of the Survey at the 
mines from which the coal was shipped and were sent to the laboratory 
in sealed cans. They were taken on opposite sides of the mine or in 
places widely separated, so that to a certain extent they show the 
variations in the quality of the coal within the limits of the mine. In 
the final report the points at which the samples were obtained will be 
accurately described, but in the present report the samples will be 
designated merely u mine sample A" and ''mine sample B." 
The laboratory car samples were taken from the carload of coal 
sent to the plant for testing at the time that it was unloaded. They 
usually represent either the whole car or that portion of it used in the 
tests. The laboratory car sample was obtained after the coal had been 
passed through rolls having an aperture of about li inches, by taking 
portions at short intervals from the buckets of the conveyor. These 
portions, after thorough mixing, were used for the preparation of the 
laboratory sample. 
LOSS ON AIR DRYING. 
Most coals rapidly lose moisture on exposure to air. In grinding 
the large sample down to the fine state of pulverization necessary in 
the small laboratory sample there is liable to be a large and undeter- 
mined loss of moisture. To reduce the error due to this loss ol 
30 
