In the preparation of the poisoning-jar it is well to use a jar which has a ground-glass stopper, and the mouth of which is about three inches in diameter. This will be large enough for most specimens. The one-pound hydrate of chloral jars, provided with glass stoppers and sold by Schering, make the neatest collecting-jars that are known to the writer. I have found it well to have such jars partly covered with leather after the fashion of a drinking-flask. An opening in the leather is left on either side, permitting an inspection of the contents of the jar. The leather protects from breakage. At the bottom of such a jar a few lumps of cyanide of potash, about the size of a filbert, should be placed. Over this may be laid a little cotton, to prevent the lumps from rattling about loosely at the bottom of the jar. Over the cotton there is pasted a sheet of strong white paper, perforated with a multitude of holes. In securing the white paper over the cyanide, the writer has resorted to a simple method which is explained in the annexed diagram. A piece of paper is placed under the jar, and a circle the size of the inside of the jar is traced upon it. Then a disk is cut out about three quarters of an inch greater in diameter than the original circle (Fig. 47). The paper is punctured over the entire surface included within the inner line, and then, with a scissors, little gashes are made from the outer circumference inward, so as to permit of the folding up of the edge of the disk. A little gum tragacanth is then applied to these upturned edges; and it is inserted into the jar and pasted securely over the cyanide by the upturned flaps. A jar thus charged will last for a long time, if kept properly closed when not in use. Cyanide of potash has a tendency to deliquesce, or melt down in the presence of moisture, and in very humid climates or damp places, if the jar is not kept well stoppered, the cyanide will quickly become semi-fluid, the paper will become moist, and specimens placed in the jar will be injured or completely ruined. It is well, however, to bear in mind the fact that the fumes of hydrocyanic acid (prussic acid), which are active in producing the death of the insect, will not be given off in sufficient volume unless there is some small amount of moisture present in the jar; and in a very dry climate the writer has found it sometimes necessary to add a drop or two of water from time to time to the cyanide. The same method which has been described for charging a jar with cyanide of potash can be employed in charging it with carbonate of ammonia. [Illustration FIG. 48.--Method of disabling a butterfly by pinching it when in the net.]