cvi GENERAL RESULTS OF THE MAKERSTOUN OBSERVATIONS. 
of the minimum in the diurnal variation ; they are also the night-hours during which the effect of the full moon 
must be greatest: upon the whole, this mode of determining the fact, from a short series of observations, seems 
open to the fewest objections. The following Table contains the means for groups of 8 or 4 days. 
TABLE 99.—Variations of the Extent of Clouded Sky for the Six Observation Hours 9" p.M. to 2" A.M., 
with reference to the Moon’s Age and Declination for the Years 1844-5. 
| After 
Moon’s Moon 
Age. farthest 
North. 
a. a. 
27— 1 
2— 5 
6— 8 
9—12 
13—15 
16—19 
20—22 
23—26 
227. The values for each year indicate that the extent of sky clouded was greatest about full moon, and least 
about new moon ; this is shewn with greatest distinctness in the means for 1845. We obtain the same result 
if we take from Table XXXIX., p. 66, the means for the 15 days with full moon in the middle, and for 15 days 
with new moon in the middle; these are, for 1844, 6-72 and 6°37; for 1845, 7:10 and 6-23 respectively. 
the mean 15 days about full moon = 6:91 
For the years 1844.and 1846, ie mean 15 days about new moon = 6°80. 
It may be a question still, how far error of estimating the extent of clouded sky in the presence and in the 
absence of the moon may enter into the production of this result. It is conceived that the effect of error in 
estimation must be nearly constant: in dark nights the extent of clouded sky was estimated by the space shew- 
ing clear stars; and it is not improbable that the extent of cloud might be rather over than under estimated 
during the absence of moon-light ; an error which could only have diminished the distinctness of the result 
obtained. Before we refer the result to the heating effect of the moon, there are other co-ordinate facts to be 
considered with reference to the motion of the atmosphere. (See No. 200.) We may inquire, however, how 
far it agrees with the heating effect of the sun, thus ;—the extent of clouded sky appears greatest in summer, 
and least in winter, it appears greatest near noon, and least near midnight; apparently, therefore, the heating 
effect is to increase the amount of cloud, and, by analogy, we should have the greatest amount of cloud about 
Full moon.* 
* Since the previous investigation was performed, I have met with a passage in Sir JoHN HERSCHEL’S very excellent “ Outlines 
“ of Astronomy,” page 261, in which he supposes that the lunar heat is extinguished in the upper regions of the atmosphere ; and adds, 
‘ Some probability is given to this by the tendency to disappearance of clouds under the full moon, a meteorological fact (for as such 
“ we think it fully entitled to rank) for which it is necessary to seek a cause, and for which no other rational explanation seems to 
offer.” He adds as a note to the parenthesis,—“ From my own observation, made quite independently of any knowledge of such 
* tendency having been observed by others. Humboldt, however, in his personal narrative, speaks of it as well known to the pilots 
“and seamen of Spanish America (H).” 
Sir Joun’s observations were probably purely qualitative not quantitative. I have much difficulty in making any objection to 
the conclusions of so accurate an observer, at the same time if his observations were not of comparative measurement, I must point 
to the previous conclusions from two years’ estimations, and add my own qualitative observation for a considerable period, that the 
clouds are both formed and dissipated under the influence of full moon, and that they are chiefly cirro-cumuli, or of that kind which 
I have termed cirro-cumulo-stratus (See No. 217), noticed frequently during the existence of the aurora borealis as the growing and 
dissipating cloud. Whether the resultant is an excess or defect of cloud during full moon, as compared with other periods, I have 
no impression, and think it extremely difficult to have any. Sir JOHN refers, in an addendum, page xv. of his “ Outlines,” to what 
he conceives a fact confirmatory of his conclusion, thus :—‘ M. Araao has shown, from a comparison of rain registered, as having 
* fallen during a long period, that a slight preponderance in respect of quantity falls near new moon over that which falls near the 
“ full. This would be a natural and necessary consequence of the preponderance of cloudless sky about the full, and forms, therefore, 
“ part and parcel of the same meteorological fact.” It will be seen, No. 235, that this result has also been obtained from the 
Makerstoun Observations, but it may still be a question whether it is confirmatory of Sir JoHn’s conclusion. When we compare the 
annual extent of sky clouded with the annual fall of rain, we do not find any direct connection: I do not know whether the diurnal 
