168 RECORDS OF DISTURBANCES 



80. Plate XXXIX deals with two occasions in September, 1902, separated by 12 days, when the H trace 

 shows in immediate succession two well-marked oscillations of similar amplitude. On September 6-7 

 there are four well-marked crests in the H curve, but the first and the last of the oscillations are much 

 smaller than the intermediate two. On this day there are also well-marked waves in the V curve, 

 troughs (minima) answering to crests (minima) in the H curve. The D curve on this occasion is also of 

 a distinctly undulatory type, but its phase seems on the whole a little in advance of that of the H curve. 



On September 18-19, 1902, the V magnet was out of action. The D curve, though exhibiting a 

 tendency to wave-like formation, is rather complex. 



Both these occasions relate to hours near midnight, when the special type of disturbance tended to 

 assume the more sharply oscillatory character usually associated with the morning hours. 



81. Plate XL shows four examples of early-morning disturbances. On June 21-22 and July 9, 1902, 

 the movements in the V curve are still of a comparatively simple type. On the former occasion we have 

 long wave-like movements in V, but towards 0.30 a.m. minor oscillations are distinctly visible. The H 

 movement in this case, though not quite symmetrical, closely resembles the special type ; it represents 

 what was unusual a considerably larger component of disturbing force than either the D or the V 

 movements. 



On July 9, 1902, the H range is again decidedly the largest. The disturbance on this day on the 

 whole resembles the special type, but a second wave has apparently rolled in before the first was 

 exhausted, so that the V trace appears double-headed. On the whole, the D and H traces are opposite in 

 phase to one another. 



January 10, 1903, and December 10, 1902, are examples of large disturbances. On January 10, 1903, 

 the H trace was beyond the limits of registration during the whole interval represented, and even the D 

 trace was off the sheet for a time. The range of D measured to the edge of the sheet was almost 4, and 

 the time from the minimum to the maximum shown was only 12 minutes. The corresponding change in 

 force is about 450y. The V trace, though showing minor oscillations, is still of a rounded wave-like form, 

 with crests (maxima) answering to troughs (maxima) in D. The next two days, January 11 and 12, 1903, 

 it may be remarked in passing, exhibited movements of similar magnitude to those shown on January 10 

 throughout the same morning hours. 



On December 10, 1902, the D trace was beyond though apparently very little beyond the lower 

 edge of the sheet from 5.7 to 5.8 a.m., and it just reached the other edge at 6.3 a.m. The most rapid 

 changes were from 4.22 to 4.36 a.m., when D increased 3 33' in 14 minutes, and from 5.8 to 5.26 a.m., 

 when it fell 3 57' in 18 minutes. In this instance there are numerous oscillations on the V trace, but, on 

 the whole, it is clear that crests (maxima) on it answer to troughs (maxima) on the D curve. The H trace 

 was beyond the limits of registration all the time. 



82. Plate XLI also illustrates morning disturbances. Only the H and V curves are shown, the D trace 

 being largely off the sheet and highly oscillatory. All three examples exhibit numerous oscillatory 

 movements, mostly very irregular. On February 8, 1 903, the H curve, however, shows a marked tendency 

 to form sharp-peaked waves, the time of the combined up-and-down movements in each occupying about 

 25 minutes. On the whole, crests (minima) in H seem to answer to crests (maxima) in V. 



Even on February 10 and March 6, 1903, though there are numerous short-period oscillations, the 

 movements in the H curve suggest as a substratum a comparatively regular wave motion of longer period. 



83. Plate XLII contains examples of very sharp oscillations of considerable size. The disturbances on 

 February 25 and 26, 1903, were recorded in the early afternoon; the others in the forenoon. Of the five 

 examples that from April 1 is the most striking. On that occasion the three elements all exhibit one 

 conspicuously large and rapid to-and-fro movement just after 9 a.m. 



Within from 6 to 7 minutes D first diminishes 84' and then increases 70', H falls and rises 90y and V 

 falls and rises lOOy. The turning-points for the three elements appear to be really identical in time. 



Individually considered, the two specimens from June 10, 1903, are much less striking, but it is a little 

 remarkable that two movements so similar in type and in magnitude should occur within an interval of 

 five hours. 



84. Plate XLIII contains examples of morning disturbances. There are here no longer rounded wave- 



