SOLAR OBSERVATORY. 145 



applied to these results. This assumption is presumably valid, for it appears 

 from Mr. Abbot's discussion of the bolometric results for 1906, that they can 

 be duplicated from pyrheliometric and relative-humidity observations with an 

 average deviation for a single observation of 1.5 per cent. 



Mr. Abbot has furnished a report on the work of 1905-06 (see pp. 152- 

 153), from which it will be seen that results of great value were obtained. 



Magnetic Observations ivith a Recording Variometer. — It does not seem 

 likely that material advantage will result from the continuance of variometer 

 records on Mount Wilson. The variometer was originally provided for the 

 purpose of noting any eruptive solar phenomena that might occur at times 

 of magnetic disturbance. No such connection has been detected, and it 

 therefore seems probable that the correlated study of magnetic and solar 

 phenomena can best be effected with the aid of records obtained at a per- 

 manent magnetic observatory, farther removed from electrical machinery 

 than is possible on the summit of Mount Wilson. 



Special Investigations. — Professor Julius's work on Mount Wilson con- 

 sisted mainly of photographic studies, with the 5-foot spectroheliograph, 

 of the anomalous dispersion of sodium vapor in an electric furnace. When 

 the pressure of the vapor in a tube heated by the furnace is reduced, and 

 the conditions so adjusted as to render the vapor as homogeneous as possible, 

 the D lines in transmitted sunlight retain their normal appearance. The 

 large tube which contains the sodium is traversed by two small longitu- 

 dinal tubes, through either of which a current of air can be forced. The 

 density gradient thus produced gives rise to strengthened dark wings of the 

 D lines, the position of which, to the red or violet, is determined by the sense 

 of the asymmetry in the illuminating sunlight. An image of the interior of 

 the tube, formed on the slit of the spectroheliograph, permits a photograph 

 to be taken, which shows the redistribution of light in the space between the 

 two small tubes, as defined by the direction and magnitude of the density 

 gradient. A series of such photographs will soon be published in one of 

 the Contributions. Asymmetrical illumination may be secured near a sun- 

 spot, from the edge of which photospheric light reaches us after transmission 

 through calcium and other vapors where density gradients doubtless exist. 

 It should be possible to learn whether the H^ flocculi are the effects of 

 anomalous dispersion by photographing a spot group with the camera-slit set 

 on H^. If photographs of the same group are taken simultaneously with the 

 same spectroheliograph, having a second camera-slit set on the continuous 

 spectrum, the distance between the edges of two spots should, in general, 

 differ in the photographs. If any differences exist, they should increase as 

 the camera-slit is set nearer H^. Furthermore, photographs made with the 

 camera-slit set on opposite sides of Hj should give differences in the form 

 and appearance of the flocculi. It is hoped that these tests can soon be made 

 with the tower telescope and 30-foot spectroheliograph. 



