1892 - 93 .] A. J. Herbertson on Atmos'phere at Ben Nevis. 181 
glaring exceptions are easily explained, and causes for some other de- 
viations may be given. Experiment 96 gave t — - 3°’ 05, t-t’ — 3° *35, 
and w=l'Q, an amount largely in excess of that given by other 
experiments for somewhat similar dry and wet bulb readings. On 
referring to the general weather notes, it was found that a slight 
drifting of snow began during the experiment. In ^o. 90, ^ = - 0°*7, 
t - t’ = 'd°A, 2 ^' = 0‘86. In this case the wind blew the smoke from 
the kitchen chimney towards the Stevenson screen. Again, the 
effect of a dense haze explains the high value, i(? = 3'88, in No. 36 
for ^ = 2°’8 and = 2°*35. Leaving out of account these three 
cases, the experiments which give results higher than those sum- 
marised by these lines are almost equal in number to those whose 
values are lower. It is worth noticing that in the latter case the 
dust records are invariably very low, and the air calm or almost so ; 
in the former case the dust particles, while never low, are scarcely 
above the average, but the wind is usually high. It seems possible 
that variations in the number of dust particles influence the results, 
but sufficient data for estimating this influence have not yet been 
obtained. 
Compared with Hygrometric Tables . — In comparing the diagram 
with one made by using similar elements obtained from Guyot’s 
Tables, it is found that the corresponding lines drawn through 
points of equal weight of water vapour per litre are approximately 
of the same sort, viz., parallel straight lines; but they cut the .^-axis 
at different temperatures from the lines in the figure. How far this 
may be due to difference of elevation cannot well be determined 
until experiments have been tried at different altitudes. 
The dew-point has' been reduced from the observed readings of 
the dry and wet bulb thermometers by means of Glaisher’s Tables, 
and the amount of water vapour in saturated air, at this temperature 
and at the summit pressure, compared with observed values. The 
observations made during the very dry weather at Christmas cannot 
be used since these tables do not give values for such low dew-points. 
It would seem that observed values are in excess of tabulated ones 
below a dry-bulb temperature of about 25° E., while they are in defect 
above it. This would agree with the results obtained by Mr H. N. 
Dickson with Professor Chrystal’s form of Mr Line’s hygrometer. 
The tables used by the different meteorological services for 
