﻿44 cox. 



minutes' preliminary treatment is small and in all subsequent experi- 

 mental tests the time of the preliminary heating was four minutes." The 

 experiments which follow will show that with some Philippine coals a 

 longer period of preliminary treatment is necessary to give very accurate 

 results. In the coals tested, the determinations of the volatile combustible 

 matter are given as ascertained by the official method and two others, 

 with are intended to avoid the quick application of heat and therefore the 

 loss which ensues in some Philippine coals. The two methods are the one 

 above described, which I have called our transition method, and another 

 1 which consists in a smoking off process; that is, one which subjects the 

 sample to a low heat, which is regulated by slowly moving a small 

 flame back and forth under the crucible, the flame being just enough to 

 keep a visible amount of smoke rising from the crucible but not sufficient 

 to cause the smoke to burn at the edges of the crucible. It is important 

 that the crucible should not be allowed to cool after the operation has 

 been begun, as in that case air would be drawn in to the coal, which would 

 cause the oxidation of a part of the fixed carbon. The most delicate 

 stage is the one when the hydrocarbons have practically all been expelled 

 and only hydrogen is still being liberated. At this point it is very 

 difficult to drive off the gas slowly enough to prevent its ignition, for the 

 smoke then no longer serves as a gauge. If the gas ignites, it is usually 

 coming off fast enough to carry with it some of the solid carbon particles, 

 as will at once be seen by the sparks ; however with care and practice this 

 can be controlled. Since the eye of the operator is the only criterion, no 

 definite time is prescribed for this preliminary treatment, but seven to 

 nine minutes are ordinarily necessary for its completion; in one extreme 

 case sixteen minutes were required. However, it is not a question of an 

 extreme amount of time but of putting the time in the right place. The 

 volatile matter should be smoksd off as fast as allowable so as not to 

 produce sparks, but not fast enough for the gases to burn. When this 

 process is completed, without disturbing the crucible, the platinum triangle 

 and crucible are quickly placed over the regulation Bunsen flame and 

 gradually lowered until they are finally in position. These conditions 

 should be maintained as nearly as practicable. 



There are times when it is very difficult to make the gas of this labo- 

 ratory conform to the requirements of the regulation flame. It has been 

 the writer's practice to use a shield to protect the flame of the Bunsen 

 burner from air currents, since the condition of the Committee on Coal 

 Analysis, that "the determination should be made in a place free from 

 drafts," is not easily attained in a laboratory in the Tropics. That this is 

 of minor importance when the regulation flame is carefully maintained is 

 shown by the following experiment. Four samples of thoroughly mixed, 

 non-coking coal were weighed out and carefully smoked off. Nos. 1 and 2 

 were finally heated for seven minutes over the full flame of a Bunsen 

 burner in a place free from drafts, while 3 and 4 were heated for the 



