THE SMITHSONIAN INSTITUTION. 217 



grws. The reading should be made at all times, and especially in 

 the winter, through the panes, and without opening the window; 

 otherwise the temperature of the chamber will inevitably influence 

 the thermometer in the open air. The degrees must be read, and 

 the fractions carefully estimated in tenths of degrees. After having 

 rapidly taken the observation, anotlicr should be made to verify it. 

 If there are several other instruments to observe, and the thermometer 

 is to be read first, the first reading may be made some minutes before 

 the hour ; the second, after the reading of the psychrometer ; and if 

 there is a difference, the mean number is to be entered in the journal. 

 When, notwithstanding the shelter, the bulb of the thermometer is 

 moistened by rain or fog, or covered with ice or snow, it is necessary 

 to wipe it rapidly, and not to record the degree until the instrument 

 has been allowed to acquire the true temperature of the air. 



Verification. — Verify the zero point, at the beginning and end of 

 winter. For this purpose, fill a vessel with snow, immerse the bulb 

 of the thermometer in the middle of it, so as to be surrounded on every 

 side by a layer of several inches of snow, slightly pressed around 

 the instrument. The stem must be placed exactly perpendicular, and 

 covered with snow as far up as the freezing-point on the scale. Let 

 it stand so for half an hour or more, and then read it, taking great 

 care to place the eye at the same height as the summit of, the mer- 

 curial column. If the top of the column does not coincide with the 

 freezing-point of the scale, the exact amount of the dififtrence must be 

 ascertained, and the correction immediately applied. At the same time 

 enter in the journal, under its appropriate head, the day on which the 

 expei iment is made, its quantity, and the moment at which the ap- 

 plication of it was commenced. [It is necessary to add that since the 

 zero point of the thermometer is not that of the temperature of snow 

 as it is frequently found when exposed to the atmosphere, but that of 

 melting snow, the experiment must be made in a place above the tem- 

 perature of freezing. Instead of snow, pounded ice may be em})loyed.] 



[Green's thermometers have an arrangement by which the tube can 

 be slipped down the small quantity necessary to correct for this 

 cliange. The end of the tube is fitted into a small plate of German 

 silver, and this fastened by a screw to the scale. If, On testing the 

 thermometer, the mercury be found to stand above 32°, free the screw- 

 one or two turns without taking it out. and push down the plate the 

 necessary amount to bring the mercury to coincide. Tlie thermo- 

 meter must be handled with great care in making this adjustment, 

 and it may be well, for additional security against accident, to loosen 

 all the screws which fasten the bands around the tube — it will then 

 slide in them more freely. After aimpleting the adjustment, they 

 may again be set moderately tight. The object of this adjustment 

 being only to avoid the trouble of making a correction, it is not ad- 

 visable to attempt it, if the observer thinks that he risks, in so doing, 

 the safety of his instrument. As the tubes of these standard ther- 

 mometers are kept for a considerable time before fixing the zero point, 

 in most cases the moving will not be required. After the first year 

 the zero point changes little, and practically, when exposed only to 

 atmospheric influences, may be considered permanent.] 



