290 LECTURES TO SCIENCE TEACHERS. 



there are less errors produced by addition than by sub- 

 traction in cases of this sort. 



Of course you will see that the calculations necessary 

 for determinations of this kind are something frightful. 

 You have to measure the height of the column of mercury, 

 the height of the barometer, to take into consideration the 

 temperature of the atmosphere, and, in very accurate ex- 

 periments, the temperature of the mercury column in the 

 barometer. This is done by means of this small thermo- 

 meter placed within it. In the case of the explosion 

 eudiometer several precautions have to be taken. One 

 which is essential is that at the bottom of the trough in 

 which the explosion takes place there must be a pad of 

 india-rubber against which the bottom of the eudiometer 

 is placed. This is a eudiometer which I constructed some 

 time ago, and which is a slight improvement on that of 

 Bunsen, in this respect, that the wires do not project 

 beyond the surface of the tube. If you have to deal often 

 with a long eudiometer of that kind, and have to clean the 

 outside of it, there is great danger of these pieces of wire 

 being caught by .the cloth and gradually bent one way or 

 the other until they are broken : and very often, previously 

 to the breaking of the wire you break the tube, from the 

 slight cracking of the glass at the outside. In this 

 one the wires have been cut off level with the surface 

 of the glass, and then smoothed by rubbing on a piece of 

 ground glass until you can scarcely feel their points. In 

 this way there is no fear of the tube being cracked by the 

 bending of the wire, and the contact is made by means of 

 an ordinary American clip with two little pieces of platinum 

 put within it to which wires are affixed, and to these you 

 can hang the wires from the coil. If they break no harm 

 is done. This tube contains a mixture of atmospheric air 

 and hydrogen, and is arranged exactly as it would be for 

 the determination of the amount of oxygen in the air. 

 After the connections have been made, the tube must be 

 held very firmly with the mouth pressed against the 

 bottom of the trough on the caoutchouc pad. The ex- 

 plosion, unless you have an improper mixture, will not 

 damage the tube ; but you should hold it at the lower 

 portion where the mercury is, because it is found that 

 wherever, such a tube bursts it is always at the surface of 



