90 



BET. G. HENSLOW ON THE 



57° to 61^ by night, so that the effects from this cause may be 

 practically put on one side. Similarly the humidity of the air ot 

 the room may be neglected. They were all carried on between 

 May and September iu the years 1880-85. 



The most disturhins element was evident] v the variable amount 

 of light or its intensity ; and this is unfortunately, as remarked, 

 just the very thing which cannot be estimated*. I have already 

 quoted AViesner's experiment, which showed how even a passing 

 cloud may affect tlie amount of vapour transpired, so that to 

 deduce mean results of any value, a very considerable num*her ot 

 comparative experiments must be made. The following examples 

 may be taken in illuistration. The specimens em{)loyed were cut 

 shoots in test-tubes as det^crlbed ; and it wdll be seen that while 

 in the case of the JEiionijmus japoniciis and Pelargonium the 

 blue liglit gives a minimum, yet this result docs not agree with 

 tliose resulting from more prolonged experiments Avith growing 

 plants, as will be seen further on, where the mean loss of several 

 days' duration is almost invariably greater under blue than 

 yellow glass ; and very rarely did the yellow glass permit of a 

 greater loss than the red. Again, clear glass usually gives 

 another maximum, as in the case of the Portugal Laurel and 

 Fern; but with the Euonymus it gave the same loss per hour 

 as the green glass, under which it is usually feeble. 



1. Experiments during Short Periods with Detached Leaves. 



Portugal Laurel. 



Duration. 



Colour. 



1 



Total loss, 

 gramme. 



Loss per 

 hour. 



Mean temp. 



F.t 









gramme. 



9.o3 A.M. to 11.34 A.M.... 



E. 



■905 



•59 



59-5 



11,34 to 2.7 P.M.... 



Y. 



•305 



•1-2 



60 



2.7 P.M. to 4.26 ...: 



G. 



•335 



•14 



HOO 



4.2*- to 5.45 



B. 



•190 



•14 



610 



5.45 to 0.35 



V. 



•120 



•14 



600 



G.35 to 7.35 



i 



CI. 



•170 



1 



•17 



59-0 



I attended a meeting this spring (1SS5) at which this subject was dis- 

 cussed by the leading meteorologists of the day. The general conclusion arrived 

 at appeared to be that no instruiueut at present invented can give more than 

 roughly approximate results of the quantity of direct sunshine recorded ; none, 

 the actual amount of sunlight, and that some instruments in use give very 

 erroneous results indeed. 



t The mean temp, is obtained from the max. and min. of the period. 



