152 G. Smeal: 



ings of the wet-bulbs had been kept clean by periodical washing 

 during the first series of experiments, but on resuming this was 

 neglected until August 10, when 31 more sets had been obtained. It 

 was then found that they were considerably soiled, and an examina- 

 tion of the results indicated that this had seriously affected the 

 readings. After cleaning, some superfluous portions were cut away, 

 leaving the bulbs somewhat freer, and thenceforward the nrnt^'rials 

 were regularly washed. The results herein detailed comprise only 

 the observations since August 10, numbering 103 ; some figures 

 obtained from the other 31 sets will be presented, as evidence of Ihe 

 necessity for clean coverings to the wet-bulbs. During the last month 

 of the work obsex'vations were mostly restricted to hot, dry days in 

 order to gain experience of low humidities. The range of tempera- 

 ture and humidity has thus been largely extended, the humidities 

 now varying between 18.01 per cent, and 95.11 per cent., and the dry 

 temperatures between 7.350 jv^d 31.350C, while the values of t-t' 

 extend up to H.GOC. 



Observations. 



The wet-bulb readings are given in Table I., with the differences 

 between them, the thermometers being numbered 2 (linen), 3 (cotton) 

 and 4 (silk), as before. 



Table I. 



