286 LECTURES TO SCIENCE TEACHERS. 



should like to say that I think the graduations on this tube 

 of Desaga's are too long. I will give you my reasons for 

 objecting to them, and then leave you to judge for yourself. 

 I have put on the board five lines, which we will suppose 

 to represent five divisions of the tube. "When those are 

 long, the interval between them appears to be less than 

 when the lines are shorter. The lines are really placed 

 about an inch apart, but they appear no doubt somewhat 

 closer together. I will diminish the length of them, and I 

 think you will find as the process of diminution goes on 

 the apparent distance between the lines increases. This is 

 really only an instance of the " scientific use of the imagina- 

 tion," but it is a very useful one. If I may be allowed to 

 direct attention to one or two of my own graduations, you 

 will find the length of the lines is only about a millimetre, 

 that is to say the length is about equal to the distance 

 between the two divisions, and I think in that way you get 

 about the maximum apparent distance, which enables you to 

 measure to about the one-tenth of a division without much 

 difficulty. 



The next process after the graduation of the instrument is 

 the calibration of it ; that is, you want to know the value of 

 each division of the whole tube. The calibration is made by 

 introducing measured volumes of mercury into the eudio- 

 meter. The eudiometer is placed in an inverted position 

 t uly vertical, being adjusted by means of a small plumb- 

 line placed at the side of it. A glass tube, closed at one 

 end, held in a wooden clamp to prevent change of tem- 

 perature, and ground carefully at the upper surface so that 

 it may be completely closed by a plate of glass, is filled 

 with mercury by means of a little reservoir, the mercury 

 being taken down to the bottom so as to prevent any 

 air-bubbles being formed on the side. This little non- 

 graduated tube is called the calibrating tube. When it is 

 completely full it is removed and the glass plate forced 

 on the top so as to expel any excess of mercury, and you 

 then have the tube absolutely filled. This quantity is now 

 carefully poured into the eudiometer which is to be 

 calibrated. If any air-bubbles remain at the side they 

 can be removed by shaking, or more advantageously by the 

 introduction of a little piece of whalebone pressed against 

 the side. After the additicn of each volume of mercury 



