'958 
Proceedings of the Royal Society 
bulb thermometers used in conjunction with Daniell’s and Regnault’s 
direct hygrometers, extending from the year 1841 to 1854. The 
difference of temperature between the wet and dry bulbs is mul- 
tiplied by the factor corresponding to the temperature of the dry 
bulb, and the result subtracted from the dry bulb temperature gives 
the temperature of the dew-point. This obviously gives us in the 
•diagram a line involving the curve connecting temperature with 
vapour pressure. Taking the wet bulb temperature as constant, and 
gradually raising that of the dry bulb, we have at first a part of the 
curve closely approximating to a straight line, and the higher the 
temperature taken for the wet bulb the closer is this approximation, 
and the longer is the part of the curve. As we continue to raise 
the temperature of the dry bulb the curve becomes less steep, and 
eventually we have a point of inflection, after which, as the dry bulb 
rises, the dew-point given also rises, until another point of inflection 
is reached. 
The only part of the curve which has to be looked at, however, 
is that first mentioned. We find this to be nearly straight, starting 
in a direction steeper than the hygrometer lines, gradually lessening 
in steepness till it crosses them, and then going on above them. 
Hence it appears that the dew-points given by these tables are at 
first lower than those found by the hygrometer, the difference first 
increasing to a maximum, and then diminishing to zero. After the 
point of coincidence the calculated dew-points are above those got 
directly, and the differences go on increasing rapidly. The first 
differences referred to, where the calculated dew-points are the lower, 
are smaller the higher the temperatures concerned, and the points at 
which the lines cross are further from the commencement of the 
curve. Hence Glaisher’s tables are practically nearly accurate until 
the point of the crossing of the lines is reached, and this is rarely 
done at the temperatures and with the humidities ordinarily ex- 
perienced at low levels in this country, although it is often passed 
at Ben He vis Observatory. 
It is to be remarked that the results deduced from Glaisher’s 
tables, used within the limits above mentioned, do not on the whole 
agree so well with direct observations at high levels as with those 
low down. In observations at Grantoff Marine Station it was found 
that in most cases the dew-point derived from both sources agreed 
