18(37.] auring two Eclipses of the Sun [in 1858 andlS67). 423 



On considering the columns of figures with attention, it will be perceived 

 that on each occasion all the thermometers in the sunshine sank as the 

 eclipse advanced, so as to reach the greatest depression not at, but after the 

 epoch of greatest obscuration, and from this point rose again as the obscu- 

 ration diminished, but in neither case arrived at the same elevation as in 

 the beginning of the eclipse. In each case the eclipse began with fair 

 prospects, but was followed by rain or snow. 



In the annular (almost total) eclipse three thermometers, two in the 

 sunshine and one in the shade, reached the very same point (47°*5), that 

 being the lowest observed ; in the partial echpse, three thermometers cor- 

 responding to the above reached nearly the same point (36°- 7, 38°, 38°), 

 the lowest observed. The lowest point was not reached on either occasion by 

 these instruments till some minutes after the moment of greatest obscura- 

 tion (6 minutes in the annular and 9 minutes in the partial eclipse) ; while 

 the thermometer enclosed in a tube did not sink below 41° at the same time. 

 The later occurrence of the extreme depression in the partial eclipse was 

 occasioned by the additional cooling influence of the clouds which gathered 

 five minutes after the epoch of greatest obscuration. 



By representing the observations in curves with ordinates proportioned 

 to the depressions at the successive epochs, the circumstances which have 

 been referred to are clearly seen, — the convergence of all the lines beyond 

 the time-point of greatest obscuration — the exactitude of this convergence 

 in the uniformly clouded sky of 1 858, and the comparative confusion of 

 the lines in the suddenly altered sky of 1867, where the effect of the access 

 of cloud is 1°'9 on the enclosed thermometer, 3°'5 on the clear exposed 

 bulb, and 5°*2 on the black exposed bulb (see Plate IX.). 



The effect of the cloud on the instruments employed in the latter half of 

 the eclipse is to reduce the temperatures at the end of the eclipse, as com- 

 pared with the beginning, more than 8° in the enclosed thermometer, 5° in 

 the dark-bulb exposed, and 2° in the clear-bulb ; but in the shaded instru- 

 ments the effect is contrary, for they gained 2°*6 between beginning and end. 



Finally, if the areas of obscuration be calculated for the several epochs of 

 observation (in 1867), and the proportions be represented by a curve 

 adapted to the scale used for temperature, the fact of the postponement of 

 the radiation-effect will appear, as well as the conformity with which the 

 temperatures follow the curve, in the bright half of the eclipse, and fall 

 away from it, but still proportionately, in the clouded half. 



EEFERENCE TO PLATE IX. 



In the diagram for 1858, the temperatures observed at the several epochs are marked 

 by the crossing of the lines S S for shade, W W for clear bulb, and B B for black bulb. 

 The central time of the eclipse is marked C C. 



In that for 1867, similar letters mark similar observations ; and, in addition, VV 

 shows the temperatures of the black bulb in vacuo, and the curve of relative obscu- 

 ration at the several epochs of observed temperature. 



