58 GREEN'S STANDARD BAROMETER. 



the amount of correction for the repaired or the new tube being 

 estimated until a good comparison can be made directly or inter- 

 mediately with the Smithsonian standard. 



The connecting the parts i andj by rings and screws, Figs. 2, 

 3, and 4, rather than by a single screw cut on the edge, is an 

 improvement, as the single wood-screw is apt, after a time, to 

 adhere so firmly that it is often difficult, and sometimes impossi- 

 ble, with safety to the parts, to separate it. 



It is not advisable to disturb the cistern, unless it becomes dif- 

 ficult, from the oxide of mercury which gradually forms, to make 

 the adjustment of the mercury to the ivory point, as there is 

 more or less risk in doing so. Any one accustomed to such me- 

 chanical affairs, with due attention to the plan, can, however, take 

 out the mercury from the cistern, refilter, clear the parts of ad- 

 hering oxide, and replace them; the instrument all the time 

 being kept vertical, with the cistern at top, as the mercury must 

 not be allowed to come from the tube. 



To insure a good vacuum by the complete expulsion of all air 

 and moisture, the boiling of the mercury in the tube is done in 

 vacn'o ; and care should be taken to preserve it in good condi- 

 tion. 



To put up the barometer for observation, suspend the baro- 

 meter by the ring A in a good light, near to and at the left side 

 of a window, and, when practicable, in a room not liable to sud- 

 den variations of temperature. Record the temperature, and 

 then, by the screw 0, lower the mercury in the cistern until the 

 surface is in the same plane with the extremity of the ivory point. 

 As this extremity of the point is the zero of the scale, it is ne- 

 cessary, at each observation, to perfect this adjustment. It is 

 perfect when the mercury just makes visible contact. If the sur- 

 face is lowered a little, it is below the point ; and if raised a 

 small amount, a distinct depression is seen around the point. 

 This depression is reduced to the least visible degree. A few 

 trials will show that this adjustment can always be made to a 

 thousandth of an inch. 



The adjustment effected, bring the lower edge of the vernier 

 C, Fig. 5, by means of the milled head D, into the same plane 

 with the convex summit of the mercury in the tube. Looking 

 through the opening, with the eye on a level with the top of the 

 mercury in the tube, when the vernier tube is too low, the light 



