30 REPORT — 1856. 



certainly very similar to what an alteration in the zero-points might do, we have 

 such evidence in special features of difference between the curves of the several 

 thermometers. Thus while ^^ and t^, by the rapid and uniform rise of their cui"ves 

 at the beginning of the period, lead one to suspect the possibility of something in- 

 strumental affecting them, yet it may be that the observations were commenced at 

 the bottom of one of the temperature waves, of which there are evidently three, with 

 a nearly sexennial period between 1838 and 1854. If this latter be the true expla- 

 nation, then inasmuch as U is retarded in its indications on t^ and ^3 by two or three 

 months, it ought to show in 1838 by so much the temperature of the opposite slope 

 of the wave, and its curve should not reach its maximum depression so pointedly in 

 1838 as those of ^4 and ^3 . On looking at it, we find t^ fulfilling these expectations 

 perfectly, for its curve, instead of rising up steeply from 1838 to 1839, is nearly 

 level. 



But there is still another proof: ti ought to exhibit the retarded effects of t^ in a 

 still greater degree, if the continued rise of ti and ^3 in 1838-39 and 1840 be due to a 

 cosmical cause, and not to an instrumental defect that would act on all the thermo- 

 meters ahke. Now t^ does precisely what it should do on such a hypothesis ; for 

 instead of being only level like t^ for 1838-39, it is even depressed, having its minimum 

 in the latter year. 



Similarly, it will be found through the whole of the period of our observations, 

 that by their regulated differences from each other depending on the effect of the 

 several depths of non-conducting matter covering each of them, the several thermo- 

 meters serve to confirm each other, as really indicating changes in the mean tempe- 

 rature of the surface of the earth, such as can hardly be attributed to any cause but 

 the variations in the development of solar light and heat. 



In this case the ascertainment of the period of the secular wave must be of the 

 utmost importance ; for its summit may bring us years warmer than any that have 

 been felt in our own day, and the bottom of it seasons with cold in corresponding 

 severity. 



On a Collimator for completing the Adjustments of Reflecting Telescopes. 

 By Professor G. Johnstone Stoney, M.A. 



This paper described an accessory to large reflecting telescopes, designed to assist 

 in adjusting their mirrors at night with more ease and accuracy than hitherto. In 

 general construction the new collimator resembles the telescopes made use of by 

 engineers ; it differs only so far that provision must be made for sufficiently illu- 

 minating the wires or an artificial star, and that its large lens should have a focal 

 length/, determined by the equation 



1— 1_1 



f~d F' 



where F is the focal length of the telescope to be adjusted, and d is the distance from 

 the centre of the large lens of the collimator to the cross wires. If this instrument 

 be placed in the usual position of the eye-pieces, the illuminated cross wires, and 

 the image of them which will be formed, may be viewed in it, and if these be now 

 brought into coincidence by the adjustment of the mirrors, the line from the inter- 

 section of the cross wires to the centre of the large lens of the collimator will be 

 the optic axis of the telescope ; i. e. this ray, after reflexion from the small mirror 

 will, if produced backward, pass through the centre of curvature of the large mirror. 

 A slight addition to the arrangement would ensure that this axis should also pass 

 approximately through the vertex of the large mirror ; but it was supposed that, 

 so far as the optical performance of the telescope is concerned, this would be found 

 a needless refinement if the collimator be employed only to complete adjustments 

 already approximately made by the usual methods, and if the small mirror be 

 properly supported. 



The experiments which had been made showed that this latter condition was one 

 of much importance and required that the support of the small mirror should be very 

 stiff, and that the small mirror should be counterpoised at the end of it. The small 

 mirror is usually supported by a single arm placed edgewise, in order to intercept 

 but little light; a second bar, also placed edgewise, aad forming a small angle 



