Notes on Barometer Construction. 49 



is finished and mounted, or on receiving some slight con- 

 cussion. We should also remember, as pertinent to this 

 question, that it is not only losing the materials and the 

 outlay of valuable time expended on the construction of 

 the instrument itself, but by the loss of such an instrument 

 after it has been brought into use, a break in the continuity of 

 our results is brought about, and we resume observations with 

 a new instrument, whose index, error, or deviation is differ- 

 ent. It would seem that with the Sprengel- pump and other 

 modern appliances at command for obtaining voids as good 

 as have been hitherto by any means obtained, boiling out 

 has become unnecessary and undesirable. 



Concerning the mounting of barometers and the mechan- 

 ical means for dividing the brass or other scales, I may state 

 that these are beyond the scope of the present notes ; but to 

 those who essay to construct for their own use this instru- 

 ment, I may mention one form of mounting which offers the 

 advantage of simplicity in the materials of construction, en- 

 listing glass and mercury only for the tube and its scale, 

 and therefore to that extent simplifying corrections of the 

 reading. On the mercurial tube mounted on a board and 

 dipping into a glass cistern there is fitted an outer glass 

 tube; the latter is divided, forming a scale which reckons 

 from a glass rod fixed on to the lower end of this outer 

 tube. This outer tube can be raised or lowered by a light 

 cord or wire passing over a small pulley, and attached 

 to a winch of o'lass rod workinof in a cork socket 

 near the mercurial cistern. Before an observation is made 

 this tube is raised or lowered until its zero pointer 

 coincides with the mercurial surface in the cistern. The 

 temperature is then taken ; the reading made and the cor- 

 rection of the column for temperature concerns merely the 

 expansibility of mercury and glass. There is a drawing of 

 this arrangement attached to a Sprengel pump in the illus- 

 tration to Mr. Mica Smith's paper on " The Motion of 

 Bodies under the Influence of Radiant Energy" in a recent 

 volume of our Transccctions. 



This completes what I have to communicate respecting 

 the selection and preparation of barometer tubes and the 

 mode of filling them, and I will therefore now proceed to 

 the description of three several proposed forms of the in- 

 strument, each of which possesses features of interest, and 

 perhaps I may correctly also state that each appears to be 



F 2 



