Fkbruary, 1903.] 



KNOWLEDGE 



3r> 



higher orders will see that the New Moon point on the 

 furve of type (2) is a point not of inflexion but of undula- 

 tion, since the tangent here does not cross the curve at 

 the point, as the tangent at a point of inflexion does. The 

 tangent here coincides with the curve in four points, not 

 in three, so that the departure of the curve from a straiglit 

 line in the neighbourhood of the point is quite insensible. 

 The path of an imaginary satellite conforming to this 

 type is shown on the diagram. 



We shall now indicate the manner in which the path of 

 the Moon, or those of the other satellites, may readily be 

 drawn. The scale should be taken as large as possible, 

 and it is impossible to give the entire curve in the case 

 of the Moon on a diagram of reasonable size ; it is, 

 however, sufficient to trace the curve from Full to 

 New, the rest of it being composed of simple repeti- 

 tions of this portion turned alternately backwards and 

 forwards. 



For a representation of the Moon's path from Full to 

 New in the pages of Knowledge, we are restricted to a 

 scale of 7 inches for 15" of the Earth's orbit. This gives 

 us as our maximum available scale one-fourteenth inch for 



Fia. .3.— The Variation Oval. 

 O = Mean Moon. • = Moon affected by variation. 



the Moon's distance from the Earth, and 28'4 inches for the 

 Earth's distance from the Sun. We therefore draw an arc 

 of a circle with this radius, and take a point E on the arc 

 as the Earth's jiosition at Full Moon. The Moon is then 

 at F on the radius produced. The time from Full to New 

 Moon is half of 29-5306 days, or 147653 days, in which 

 time the Earth moves through 14° 33i'. Her perpendicular 

 distance from the initial radius is then 7 12 inches, which 

 determines E" her position at New Moon. The Moon is 

 then at N on the radius through E". The Moon's path at 

 Full and New Moon approximates to small circular arcs 

 with radii 208 and 489 inches respectively, these corre- 

 sponding to the radii found above. Further, at last Quarter 

 we place the Moon at L on the Earth's orbit, one-fourteenth 

 inch ahead of E', the middle point of the ai-c EE". 

 For convenience of subdivision the arc EE" is divided into 

 12 portions, each described in about Ij days. The dots 

 representing the Moon are placed at a constant distance 

 from those representing the Earth, and revolving round 

 it with imiform angular velocity. The dotted curve 

 representing the Moon's motion round the Sun can 

 then be drawn, and it will he seen that it is concave to 

 the Sun at N, though much more nearly straight here 

 than at F. 



Hcttcr. 



[The Editors do not hold themselves responsible for the opinions 

 or statements of correspondents.] 



THE SOLAR DISTURBANCES OF SEPTEMBER, 

 OCTOBER AND NOVEMBER, 1902. 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — From a study of the great solar outburst of last 

 September and the subsequent disturbances of October 

 and November, I have been led to observe certain curious 

 coincidences— to use no stronger term — in regard to the 

 position of some of these disturbances and the order of 

 their appearance on the solar disc. Thus the September 

 outburst, which consisted of a chain of disturbances lying 

 pi'actically along a meridian, and' extending from a high 

 south latitude into the equatorial zone of the northern 

 hemisphere, was followed early in October by a group of 

 spots in the nor/hern hemisphere and at a distance in 

 longitude of about 180". Now allowing an apparent or 

 synodic rotation period of 26J days to the latter gro\ip 

 and 27 days to the most important portion of the former, 

 situated about south latitude 24", we find that October 

 and November gave us repetitions of the above pheno- 

 mena in a portion of the disturbed area, and that each 

 recrudescence appeared to affect a portion of the opposite 

 hemisphere situated at the extremity of a diameter. The 

 same rule also held good in regard to a group of facula* 

 independent of the above, and which, appearing toward 

 the end of October in the southern hemisphere, had its 

 vis-d-i'is in November in the northern hemisphere. 



If these facts be something more than mere coinci- 

 dences, do they not suggest the conclusion that in the 

 displays of the last three months we were viewing some 

 of the effects of eruptions ileeply seated and powerful 

 enough to shake the whole body of the sun, and to cause 

 great waves to converge at their antipodes, and to excite 

 other eruptions in turn ? 



Moneybroom, Lisburn, John M. Harc 



16th December, 1902. 



Astronomical. — In his work on the nebuliB, Sir William 

 Herschel recorded the positions of fifty-two supj)0sed 

 extensive regions of diffused nebulosity. Recognising the 

 great importance of these in considerations relating to the 

 structure of the universe. Dr. Isaac Robei-ts has made a 

 careful photographic investigation of the regions in 

 question, with the remarkable result that in only four of 

 them has any nebulosity been found, although the photo- 

 graphs show stars as faint as 17th magnitude. One of 

 these four, branching out from s Orionis, is of peculiar 

 interest. 



The interesting suggestion made by Sir David Gill a 

 few mouths ago, to the effect that the brighter stars, as a 

 whole, appear to have a motion of rotation with respect to 

 the fainter stars, has been tested at Oxford by Prof. 

 Turner. A relative motion similar to that described by 

 Sir David Gill, anil equal in magnitude, was indicat<?d by 



