Take a 2-ounce quinine bottle, or still better a shorter bottle with a wide mouth; break up a quantity of cyanide of potassium into pieces of convenient size (about a cubic centimeter); put these pieces in the bottle so that they form an even layer at the bottom; mix in a convenient vessel a quantity of plaster of Paris with water just sufficient to make the mixture semifluid and then pour it over the cyanide so as to cover this last to a depth of about 5 millimeters. The bottle is then left open for an hour or two until the plaster is thoroughly dry. The walls of the bottle are then cleansed from particles of the plaster which may have splashed on them, and the bottle is ready for use. If not used too frequently, especially in warm weather, it will last for an entire year or longer. Bottles or vials of different sizes can be prepared in the same way, and a very small cyanide vial which can be carried in the vest pocket will be found most convenient for use on all occasions. Fig. 92 represents a medium-sized chemist's test tube, converted into a very convenient cyanide bottle, in which, however, a cotton wad has been used to keep the poison in place. When the collected specimens have been removed from the bottle the latter should be carefully wiped clean with a piece of cloth or paper. The surface of the plaster soon becomes dirty and, on account of the hygroscopic property of the cyanide, more or less moist, especially during warm weather. The cyanide bottle is, therefore, not well adapted for the killing and temporary preservation of small and delicate specimens. This difficulty can be altogether obviated by placing a circular piece of blotting paper, cut to neatly fit the interior of the bottle, on the surface of the plaster. This can be renewed once a week or so, or oftener if it becomes necessary. It will frequently be advisable, also, especially in the collection of Diptera, Hymenoptera, and other delicate insects, to put a strip of blotting paper partially round the inner side of the bottle. This will absorb any moisture which may gather on the inside of the bottle and which would otherwise wet and injure the specimens. The accompanying figure (Fig. 93) illustrates a bottle arranged as described above. A similar result is attained by some collectors by partially filling the bottle with narrow strips of bibulous paper to support and separate the insects as shown at Fig. 91. [Illustration: FIG. 92.--Pocket cyanide bottle.]