<50 Transactions of tJie South African Pliilosopliical Society. 



inch thick, of different colours, placed horizontally about one inch' 

 iibove the rims of the mugs, and large enough to prevent any sun- 

 light reaching the water without having first passed through the 

 glass. The sides of the mugs were also protected by bricks from the 

 rays of the sun. The measures were taken with a graduated glass 

 belonging to an 8-inch rain-gauge, the measures being multiplied 

 by 3*55 to give the true loss in inches. The results were also verified 

 by weighing. The temperatures of the waters were determined 

 nearly every day at XVII. and sometimes at XIV. by means of a 

 ■standard Kew-certified thermometer. The following were the results 

 for the period July 20 to November 22, 1896—125 days :— 



1. July 20 to Aufjiist 30, 1896. 



Colour. 



Mean Temp, at XIV. 



Mean Temp, at XVII. 



Total Evap. 













Inches. 



Eed 



72-8 



65-4 



601 



Yellow 



72-3 



64-8 



619 



Green 



B9-6 



64-1 



5-59 



Blue 



70-8 



64-3 



6-17 



-Purplt 



72 2 



64-8 



6-27 



2. AiujHst 31 to Noveniher 7, 1896. 



Colour. 



Mean Temp, at XVII. 



Total Evap. 



Bed 



Yellow 



o 



77-7 

 77-8 

 77-0 

 77-8 



77-6 



Inches. 

 20-21 

 'JO -12 



Green 



18-76 



Blue 



19-97 



Purple 



20-38 



3. November 8 to 22, 1896. 



Colour. 



Mean Temp, at XVII. 



Total Evap. 



Bed 



Yellow 



o 



81-8 

 SI -6 

 81-1 

 82-0 

 81 -.5 



Inches. 

 4-29 

 4-22 



Green 



4-14 



Blue 



4-0o 



I'nrple 



4-30 



