212 Messrs. Warren De La Rue, Stewart, and Loewy [Mar. 14, 



had in 1866. We had then only the groups observed by Carrington, the 

 positions and areas of all of which we had accurately measured. We have 

 now in addition five years of the Kew observations, for each group of 

 which the positions and areas have been recorded by us in our previous 

 communications to this Society. We have thus altogether observations 

 extending from the beginning of 1854 to the end of 1860, forming the 

 series of Carrington ; and observations extending from the beginning of 

 1862 to the end of 1866, forming the Kew series, as far as this is yet reduced. 

 We have, in fact, altogether a nearly continuous series, beginning a year or 

 two before one minimum, and extending to the next, and thus embracing 

 rather more than a whole period. 



We propose in the following pages to discuss the behaviour with regard 

 to size of the various groups of these two series, as each group passes from 

 left to right across the sun's visible disk. Unfortunately for this purpose, 

 a large number of groups has to be rejected ; for, on account of bad weather, 

 we have frequent blank days, during which the sun cannot be seen, and on 

 this account we cannot tell with sufficient accuracy the behaviour of many 

 groups as they pass across the disk. In our catalogue of sun-spot beha- 

 viour we have only retained those groups for which, making the times 

 abscissae, and the areas ordinates, we had sufficiently frequent observations 

 to enable us to construct a reasonably accurate curve exhibiting the area 

 of the group for each point of its passage across the disk. From these 

 curves a Table was then formed denoting the probable area of each non- 

 rejected group at the following heliographic longitudes (that of the visible 

 centre of the disk being reckoned as zero) : — 



-63 -49 -35 -21 -7 +7 + 21 + 35 + 49 +63°; 

 in fact giving the area of the group for the ten central days of its progress, 

 and rejecting those observations that were too near the sun's border on 

 either side, on account of the uncertainty of measurement of such observa- 

 tions. We have succeeded in tabulating in this manner 421 groups of 

 Carrington's series, and 373 groups of the Kew series up to the end of 1866, 

 in all 794 groups. In this catalogue the area is that of the whole spot, in- 

 cluding umbra and penumbra ; and in measuring these areas a correction 

 for foreshortening has always been made, as described in a paper which we 

 presented to this Society, and which constitutes the first series of our re- 

 searches. These areas are expressed in millionths of the sun's visible 

 hemisphere. 



4. When we began this present investigation into the behaviour of spots, 

 we soon found reason to conclude that in the case of sun-spots the usual 

 formula for foreshortening is not strictly correct. Perhaps if a sun-spot 

 were strictly a surface-phenomenon, the usual formula might be correct, 

 though even that is doubtful ; for the earth as a planet may not impossibly 

 affect the behaviour of all spots as they cross the disk, so as to render the 

 f formula somewhat inexact. However this may be, a spot is probably 

 always surrounded more or less by faculous matter, forming in many cases 



