37 
but highly interesting branch of enquiry, and to the necessity 
of devising and adopting a more reliable and systematic 
method of determining the intensity of solar radiation than 
the one at present in use. It is much to be regretted that 
no regular and long-continued series of observations has 
ever been made with Sir John Herschefs actinometer, since 
there can hardly be any doubt that such a series would 
yield much more accurate and reliable results than can be 
obtained by the use of either the ordinary or the vacuo 
black-bulb thermometer. 
The Oxford solar radiation observations commenced on 
the 2nd of January, 1856, and were made with a thermo- 
meter supplied by Negretti and Zambra. On the 3rd of 
November, 1858, this thermometer was unfortunately 
broken, but a new thermometer being obtained, the observa- 
tions were resumed on the 1st of December, 1858, and 
continued uninterruptedly till the 11th of September, 1864, 
when the second thermometer was also broken; a third 
thermometer was obtained and brought into use on the 17th 
of September, and the observations continued with it to the 
end of the year. The entire series of observations during 
the nine years 1856-1864, must therefore be regarded as 
consisting of two distinct series, made with different thermo- 
meters, and therefore not strictly comparable with each 
other. I have, in consequence, thought it desirable to con- 
fine the discussion, in the first instance, to the observations 
made with the second thermometer during the years 1859- 
1864, more especially as in 1857 the readings of the first 
thermometer were only to the nearest whole degree, and in 
1858 the month of November was omitted; and in order 
to make six complete years, I have included the observations 
made with the third thermometer from the 17th of September 
to the end of 1864. 
The following table shows the mean monthly and annual 
differences between the maximum temperature in the shade 
