PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. V. 



629 



At 7 35, at <?, and dim. 



7 38, do. do. 



7 39, at /;, and bright. 



7 43, do. do. 



7 44, at a, and certainly dimmed. 



7 45, at b, and bright. 



7 47, at a, and dim. 



7 48.1, do. do. 



7 49, brightening. 



7 50, at b, and bright. 



7 51, at 6, and quite bright. 



7 52], dimming. 



7 52^, at rt, and dim. 



7 53-i> brightening. 



/;. in. 

 At 7 54.jr, at b, and bright. 



7 55j> at ^i ar| d quite bright. 



7 57?> at fl > ar| d quite dim, as if dying away. 



7 58} do. 



do. 



do. 



7 58! , brightening. 



7 59.}, at A, and bright. 



8 o, do. and quite bright. 



8 3, brighter than at any time yet, and has clear- 

 ly ascended to the Milky Way by lines d d. 

 8 \\, dimmed and sunk to b. 

 8 7, brightening. 

 8 8, very bright, and at dd. 



8 15, still as last, and seems to be permanent now. 



9 30, boundaries to x. 



I think I can know when it is going to be permanent, by the upper portion of the Light brightening more 

 than at any time previously in the evening, and the strong brightness ascending higher. The first 

 appearance of the Zodiacal Light seems to be a white light ;'. r. when the twilight has not quite gone; 

 afterwards it changes to a warm yellowish light. The reverse of this happens in the morning. The 

 Diffuse Light is now very dim; in the morning it is very strong, for it. 



This evening was remarkably fine for observations, and in my notes is the remark: "It certainly 

 does pulsate". 



123. Only one abruptly-beginning magnetic disturbance occurred in the period when we were ob- 

 serving with "quick-run" registrations in Assouan, namely, the gth April. 



I have, unfortunately, not received any intimation of quick-run registrations having been taken 

 except in Samoa, where Prof. Dr. ANGENHEISTER commenced the registrations on April 10, /'. e. one 

 day too late. Mr. TITTMA.NN, superintendent of the U. S. Coast and Geodetic Survey, has been good 

 enough to send me some copies of slow-run registerings for April 9 from Cheltenham, Porto Rico, 

 Tucson, Sitka and Honolulu. Of these, the curves from Honolulu (158 W) and Porto Rico (65 W) 

 are of special interest, because these stations, together with Assouan (33 E) form a particularly happy 

 distribution of stations about the Earth. 



Figure 228 shows that on this day an equatorial perturbation occurred, the character of which is 

 very similar at the three stations. The times of commencement in H are as follows: 



Honolulu Porto Rico Assouan 



loh 2o', 7 p. m. Gr. M. T. 20>, 8 20'" 44* 



The changes in D at the same time were very small, as might be expected would be the case 

 with this kind of perturbation. 



The first notices of time are given in a letter from the Coast and Geodetic Survey, - - the last 

 value is found by the "quick-run" magnetograms from Assouan as shown by the magnetogram. 



The last time-determination is given in seconds, because of the greater accuracy that can be 

 reckoned upon in "quick-run" registerings. 



The time-marks here, which refer to the central point for the obliterated parts, are certainly cor- 

 rect to one second, but a greater uncertainty arises when it is a question of determining when the 



Birkeland. The Norwegian Aurora Polaris Expedition, 1902 1903. 80 



