REPORT or THE KEW COMMITTEE. lix 



being also below 32°, cold water should be poured over the wet bulbs and 

 the connecting strings. In a few minutes the wet bulbs will by this means 

 be covered with a fresh coating of ice ; this should be repeated if necessary. 

 If this operation is performed two or three times a day during very cold 

 weather, there is reason to believe that the Avet bulb will always be covered 

 with a sufiScient coating of ice. But if the wet bulb and the water of the 

 water-vessel be frozen from previous cold, the present temperature being 

 above 32°, the ice of the water- vessel may be thawed by warm water, using 

 no more than is necessary for the purpose. 



If these regulations be followed during the cold months of the year, it is 

 believed that there are comparatively few instances where we may not know 

 the temperature of evaporation during frost. 



During diy weather the wet-bulb arrangement is again liable to go wrong, 

 although from a different cause. Tlie thread, which in the arrangement 

 adopted lies along a copper groove, gets dry in its passage from the water- 

 vessel to the bulb, the capUlary action ceasing. Sometimes it apparently 

 rights itself without aid, but sometimes it continues wrong until it is put 

 right at the next observation hour. The commencement and termination of 

 Buch a wrong state of the wet bulb are generally so clearly indicated by the 

 curve itself, that there appears to be little or no uncertainty in ascertaining 

 what observations ought to be rejected. This action would best appear to be 

 prevented by the use of an india-rubber tube lying along the metallic groove, 

 and having one end dipping into the water of the water-vessel ; and through 

 this tube the thread ought to be carried in its passage from the water-vessel 

 to the thermometer. Evaporation is thus avoided, and the arrangement 

 ■will probably answer in winter. When the supply of water is too rapid, it 

 may be easily and safely altered by turning up the tube. 



Even when the action of the wet bulb is unexceptionable, water must fre- 

 quently be added to the water- vessel. It is usual for this water to have the 

 temperature of the air ; but in cases of a great difference between the two 

 bidbs, this will be much above the temperature of evaporation; the con- 

 sequence is found to be, that in such cases there is a rise in the wet-bulb 

 curve which, in extreme cases, may not completely right itself until a quarter 

 of an hour has elapsed. This can only be remedied by each observatory 

 doing all in its power to ensure that under such circumstances the water 

 supplied to the water-vessel shall represent as nearly as possible the tem- 

 perature of the wet bulb at that moment, and also that the supply of water 

 from the water-vessel • to the wet bulb shall be no greater than is necessary 

 to keep the bulb thoroughly damp without dripping. 



With regard to other deficiencies, it will only be necessary to remark here 

 such as are peculiar to the Thermograph, since all those common to this in- 

 strument and the Barograph have already been stated under the head of the 

 latter. 



In the first place, it should be noticed that there is sufficient light to illu- 

 minate the whole range of the curve in a proper manner. In order to ensure 

 this, and at the same time procure the best possible definition, the heights of 

 the thermometers may, as occasion requires, aud without detriment to the 

 instrument, be altered so as to bring the mean temperature of the time into 

 a central position with respect to the lens and light. This change ouoht, 

 however, to be made as seldom as possible (perhaps twice or thrice in a year), 

 and when made great care ought to be taken that there is no strain upon the 

 wet-bidb Thermometer through tightness of the thread, whether arising 

 from frost or any other cause. 



