46 SCIENCE PROGRESS 
4. PREPARATION OF MATERIAL 
In the case of sands, the separation can usually be made 
without crushing, as the grains are generally small and of a 
simple character; but it is often convenient to sift, in order, 
as far as it is practicable, to secure uniformity of size among 
the grains. 
With rocks, the degree of crushing required depends upon 
the texture, but as a rule the diameter of the grains obtained 
should not exceed 0'25 mm. The crushing should be done so 
as to secure a minimum of powder, especially in quantitative 
operations. This is effected by sifting at frequent intervals 
during the crushing process. The powder should be washed 
off, and, if necessary, treated separately in the wet way according 
to Pebal’s suggestion. In any case the treatment of the powder 
usually gives unsatisfactory results, although for quantitative 
purposes it obviously cannot be neglected. On this point it is 
important to note that some minerals powder more readily than 
others, hence the proportion of any mineral present in the 
powdered portion will differ from that of the same mineral in 
the coarser portion of the crushings. 
For the purposes of isolation, or in merely qualitative 
separation, the powder may be washed off and rejected after a 
cursory examination with the microscope. 
5. SEPARATION OF THE MINERALS 
Magnetite, if present, should first be extracted by a weak 
permanent magnet—an operation which is facilitated by covering 
the poles with a movable paper cap. The portion thus extracted 
should be carefully examined, as it may contain pyrrhotite and 
even hematite, both of which are very strongly magnetic. 
Ilmenite is not extracted by a weak magnet, except in the finest 
powder, and then only to a small extent. 
The remainder of the grains are placed on cardboard, and, 
assuming the current to be switched on, are presented to the 
poles of the electromagnet. The cardboard is then removed, the 
current shut off, and the adhering grains allowed to drop into 
a cardboard box placed under the poles. This is continued till 
no more grains adhere to the poles. A certain amount of non- 
