368 Total Solar Eclipse of 1914 [Feb. 26, 



first, because the edges of the streamers are not so well-defined as in 

 the first case, and secondly, because some of the angles of the hexagons 

 are very acute, so that tiie point 0, representing the spot, could be 

 shifted, within limits, without altering sensibly the collinearity of 

 the points of intersection of the opposite sides of the hexagon. The 

 two projected conies are hyperbola. 



We now turn to the luminous sheaf of rays in the N.E. quadrant, 

 of which two bright ribs were selected for measurement. In these 

 cases the solution is again not so rigid, even as in the last two cases, 

 and the position of could be considerably varied, without sensibly 

 affecting the collinearity of the points of intersection of the opposite 

 sides of the hexagon. The reason is that these rays point almost 

 directly, with a very slight curvature, at the spot, which at first sight 

 seems paradoxical. All we can affirm is, that the six points in each 

 case lie on conies which probably pass through the region of the spot. 

 The projections being nearly straight lines would indicate that, 

 the conies of projection from the region of the spot are seen nearly 

 edgewise. 



In the N.W. quadrant, therefore, curve I (fig. 2) is an elongated 

 ellipse, which certainly passes through the immediate region of the 

 spot. Curves II and III are hyperbolae which most probably pass 

 through the immediate region of the spot. They almost certainly 

 pass through such a region as could be represented by the accompany- 

 ing faculae and flocculi. All we can affirm about curves IT and Y is 

 the probability that they pass through the region of disturbance. 

 The total result of this investigation is that the area of disturbance 

 marked by the sun-spot as a centre was the seat of emanation of 

 streams of matter, represented certainly by the projected conic of 

 curve I, very probably by those of curves II and III, and probably 

 by those of curves IV and V. From this result, and those established 

 in former investigations for the eclipses of 1893, 1898, 1905, and 1908, 

 it would follow that the mode of propagation of matter from a centre 

 of disturbance on the sun, which may be supposed to be operative 

 in the production of terrestrial magnetic storms, is in the form of 

 a set of diverging rays, and not in that of a single cathode-like 

 dischartre. 



[A.L.C] 



