EFFECT OF THE FROSTS OF 1908-9 ON VEGETATION. 363 



The lack of uniformity in making temperature records is commented 

 upon by Mr. Curtis in a note which we append, and this lack is greatly 

 to be regretted, as it makes exact comparisons between different locali- 

 ties impossible, and accounts for some of the apparent differences in 

 behaviour noted in the returns. 



Another factor making comparisons difiicult is the very considerable 

 range of temperatures found within quite small areas at different eleva- 

 tions, sometimes amounting to several degrees. This last factor it is 

 practically impossible to eliminate in returns of this kind, though the 

 former disturbing factor is capable of remedy. 



Mr. Curtis says : — 



" The value of temperature observations in such a discussion as the 

 present depends entirely upon their being so made that they shall be 

 perfectly comparable with each other, otherwise it may very well appear 

 that in one locality plants have been able to withstand a temperature 

 considerably below that at which similar plants have succumbed in 

 another, and in such a case the a priori explanation would be not that 

 the instruments had been incorrectly read, but that from some cause 

 the temperature shown by the thermometer at one of the two places 

 was erroneous. 



" The observations supplied in response to the circular in the 

 present case emphasize this point and show the desirability of using 

 verified thermometers of standard type, and also the necessity for ex- 

 posing them under similar conditions. In some instances the observa- 

 . tions were made from instruments exposed in ' screens,' but the kind 

 i of screen is not stated ; in others the thermometer was exposed upon a 

 ' post; and in yet other instances upon a wall, and in each of these 

 j exposures there are several factors such as height above ground, the 

 I character of the wall, and especially the aspect, which would materially 

 i influence the indications of the instrument. A thermometer placed 

 I four feet above the ground in a double screen may not — and, indeed, 

 frequently does not — indicate the temperature which an instrument 

 freely exposed to terrestrial radiation w^ould show, or that to which 

 , vegetation would be subject close to the ground; but the observa- 

 : tions v/ould be comparable with each other, and would enable a far 

 Ijuster view to be obtained of the distribution of cold or heat over a 

 district than could be got under other circumstances. The description 

 of the instrument which was used is also in several cases vague, and 

 the term ' ordinary ' applied to the thermometer may convey a different 

 meaning to different people. " 



* * H« H< * * * 



