148 REPORT—1840. 
instead of elaborate formule of interpolation, a method which, 
however valuable in itself, has been carried to excess by the 
German meteorologists ; and the immense labour they have 
spent in computing arbitrary constants might, in many cases, 
have been better bestowed. Anything, however, is better than 
no reduction at all. The engraving of such charts is well worth 
doing ; they often afford a check upon the accuracy of the ob- 
servation, the value of the method, and the consistency of the 
results, which no other mode of exhibition can possibly do. 
330. I would notice the projection of observations on sub- 
terranean temperature (96), as a signal instance of all the 
advantages which I have mentioned. Where several simul- 
taneous but independent observations are made, the comparison, 
by projection, gives a moral certainty of the fidelity of the ob- 
servations, whilst the coincidence of one year’s observations 
with another demonstrates the confidence due to the method of 
investigation even more satisfactorily than the coincidence of 
final results. Even in point of accuracy, the maxima and 
minima cannot be so well determined by a parabolic interpola- 
tion (as M. Quetelet has done), because the temperature -curves 
invariably rise faster than they full; consequently, the oscu- 
lating parabola has not its axis vertical, but inclined to the left 
upwards ; and if a series of circular functions be used to ex- 
press the observations, the computation becomes very laborious, 
and perhaps the result is not more accurate than may be ob- 
tained in a few minutes in the following way.—A thin stiff 
wire may, by a little practice, be adapted so as to follow, in 
a remarkable manner, the sinuosities of the temperature-curves : 
by changing its form by pinching, then bending it elastically, 
and varying the inclination of its symmetric axis with the verti- 
cal, it is easy to lay it so over the curve*, as to satisfy the eye 
that the inequalities on each side compensate one another. 
331. Until national observatories shall be formed, it seems of 
great consequence that the Plymouth Hourly Observations, 
made at the expense of the Association, should by no means 
be discontinued ; that the results, amply reduced, perhaps the 
detailed observations, should be printed at their expense ; and 
most particularly, that the condition of the instruments shall 
be from time to time verified with the utmost care, in order to 
render these determinations positively as well as relatively ac- 
curate, for which I do not know whether there is at present 
any sufficient security. 
* Or polygon; for in all cases, the actual points of observation should be joined 
by straight lines. This suggestion I owe to Mr. Babbage, and I have found it 
attended with great advantage. 
