228 THE ENTOMOLOGIST. 



the first tube above the cork outside. The end in the bottle 

 is filed or ground flat, and then, with sealing-wax, fastened 

 on to a small plate of metal (tin or zinc) ot the exact size 

 and shape shown in figure 2. This piece of zinc is bored 



with a hole one-eighth of an inch 

 Fig. 2. in diameter, at a quarter of an inch 



from one end (as shown in the 

 woodcut) ; the edges of the hole 

 must then be smoothed down per- 

 fectly by rubbing the metal up and 

 down a hone or piece of slate a few 

 limes. It is exactly over this hole 

 the glass tube is fastened, so that 

 there will be a free passage for the 

 air through both, the junction of the tube and metal 

 being rendered air-tight by the sealing-wax. Now fasten 

 by its narrow end (in the position shown by the dotted 

 line in the woodcut) a slip of thin gutta-percha tissue, 

 or, still better, of oil-silk, by means of any cement, so 

 that the broad end shall freely play over the hole ; a 

 valve opening inwards is thus made. To test the valve, 

 suck at the free end of the tube, and if no air passes the 

 valve is good; if any can be sucked through, it is useless, 

 and a new one must be made. The causes of failure would 

 be either using too stiff a piece of gutta-percha or oil-silk, or 

 pieces that have fine holes in them, or else the valve-hole in 

 the metal has not its edges smoothly ground down. By 

 attending to tliese details anyone can soon construct a valve. 

 The reason for using the india-rubber cork is that an 

 ordinary cork (unless very sound) is not air-tight to the 

 pressure of air employed, while if rendered impervious by 

 the use of sealing-wax it could not be removed to clean or 

 renew the valve. These corks, already bored, may be had at 

 any chemical apparatus manufacturers,* and are most 

 convenient, as they can be removed at pleasure. If the cork 

 be now placed in the bottle with its tubes fitted in, and if 

 the delivery tube be stopped up and air be forced in by the 

 mouth through the valve, it will remain in the bottle if the 

 valve be sound, rushing out with a suiall explosion only on 

 opening the delivery tube. Strong though the puff of air 

 * E.g., Messrs. Mercer and Townseud, Bishopsgate Street, London. 



