May 26, 192 1] 



NATURE 



393 



previous experience, which shows that, as a rule, 

 when a "sudden commencement " occurs not far from 

 noon of local time the principal maxima and minima 

 of the storm occur within the next twelve or fifteen 

 hours, but that if the "sudden commencement" 

 occurs late in the day the full development of the dis- 

 turbance is postponed until the post-meridiem hours 

 of the following day. 



The changes in the horizontal components of force 

 throughout the disturbance were on a very large scale, 

 and took place with a rapidity so great that the photo- 

 graphic traces are in some parts too faint to read. 

 But the most unusual feature of the storm was the 

 remarkable series of changes in the value of the 

 vertical component of force. In most magnetic storms 

 the general course of events comprises (i) a gradual 

 increase in V to a maximum (in some cases two suc- 

 cessive maxima) reached about i8h. local time, (2) a 

 gradual fall until midnight, (3) a rapid fall to a mini- 

 mum which is reached about ih., (4) a gradual 

 recovery to nearly the undisturbed value, which 

 recovery is completed by about 8h., and is sometimes 

 accompanied by (5) a series of short-period pulsations. 

 The whole sequence is frequently repeated on a 

 modified scale later on in the second day. In the 

 storms now described this course of events was fol- 

 lowed so far as the first twenty-four hours are con- 

 cerned. The first maximum on May 13 was reached 

 at 2oh. 37m., and was 1527 above the undisturbed 

 value at the time of the "sudden commencement." 

 One or two falls in value succeeded until 2ih. 24m., 

 when a rapid fall of more than 3507' in six minutes 

 carried the light spot off the paper. The principal 

 minimum which then occurred took place at an un- 

 usually early hour (see " British Meteorological and 

 Magnetic Yearbook," 1915, part iv., p. 89, and plate 

 vi.). According to the theory which would account 

 for magnetic storm phenomena by assigning them to 

 the earth's rotation in a flbam of particles emanating 

 from the sun, this sudden drop in the value of V 

 might be supposed to take place at or soon after 

 midnight of local time, but not so early as 9 p.m. 

 The^ rate of change in V during the descent to the 

 minimum referred to was large, but by no means the 

 largest recorded at Eskdalemuir. For example, the 

 storm of March 23, iq2o, showed a case of V chang- 

 ing at the rate of j6qy per minute. The gradual 

 recovery which followed the minimum was accom- 

 panied, particularlv between 6h. and 8h., by pulsa- 

 tions of about four minutes' period and of amplitude 

 averaging about 47. It is, perhaps, unsafe to 

 generalise, but there is some evidence to show that 

 such pulsations in vertical force do not occur in a 

 magnetic storm unless the total range of disturbance 

 in V exceeds a certain amount. 



The second twentv-four hours of the storm showed 

 even more remarkable developments than the first. 

 After several maxima and minima had been passed 

 the value of V began to fall about 23h. on May 14, 

 and the light spot went off the paper at midnight. 

 Between ih. and 6h. on Mav 15 at least a dozen 

 extensive and rapid changes in V took place, swing- 

 ing the light spot alternately beyond the upper and 

 below the lower edge of the paper. The most clearly 

 marked of these occurred between 2h. 40m. and 

 2h. 44m. on Mav 15, and involved a change during 

 that interval at the rate of 1387 per minute. Repeated 

 oscillations of this character and magnitude have not 

 hitherto been recorded at Eskdalemuir. The course of 

 the disturbance during May 16 was remarkable in 

 that the fall in the value of "V during the early hours 

 of the morning continued until nearlv yh. 



The storm had practically died down by noon on 

 May 17, but soon after s-jh. on that day another 



NO. 2691, VOL. 107] 



"sudden commencement" was recorded, and the 

 sheet which will be taken off the recording drums 

 to-morrow morning will probably show a recur- 

 rence of disturbance. 



A. Crichton Mitchell. 

 Eskdalemuir Observatory, May 18. 



Ocean Tides. 



The article in Nature of March lo, p. 33, on 

 oceanographic problems by "J. J." prompts one 

 interested in tides to direct attention to the services 

 which a new Challenger expedition might render to 

 the general theory of ocean tides at relatively little 

 cost either in time or in money. 



There are numerous localities for which tidal 

 information is either inadequate or wholly lacking. 

 Sir George Darwin directed attention to some of these 

 places by publishing in the Geographical Journal of 

 November, 1909, a memorandum prepared by the late 

 Dr. R. A. Harris, of the U.S. Coast and Geodetic 

 Survey. The "dozen or so landing parties placed 

 here and there over the world " suggested by "J. J." 

 could undoubtedly use to advantage as their bases of 

 operations some of these places for which tidal ob- 

 servations are needed. These suggested landing 

 parties could therefore, in connection with their other 

 observations, secure tidal data of great value. 



It may not, perhaps, be amiss here to point out that 

 such tidal observations would serve two important 

 purposes. In the first place, they would increase our 

 geographical knowledge of the regional distribution 

 and local characteristics of the tides ; and in the 

 second, they would furnish further data of an accurate 

 character to test the merits of the various tidal theories 

 that attempt to interpret mathematically the terrestrial 

 phenomena of the tides. Thus some of the places for 

 which tidal information is desired are of critical im- 

 portance to the so-called "stationary wave " theory of 

 tides, which appears steadily to be gaining in favour. 



The use of automatic or self-recording tide gauges 

 would, of course, be most desirable. In this connec- 

 tion it is to be noted that such tide gauges may now 

 be had in small and inexpensive types that require no 

 elaborate installation and may be expeditiously set 

 up. It cannot, however, be too strongly emphasised 

 that valuable additions to our knowledge of the tides 

 at out-of-the-way or infrequently visited places may be 

 secured by recording hourly the height of the tide as 

 registered on a naked tide-staflf graduated to feet and 

 tenths. The longer the series of observations, the 

 better ; but even a day or two will furnish considerable 

 information. 



The value of the tidal observations would be greatly 

 enhanced if bench-marks of a permanent character 

 ■were established and the relation of the zero of the 

 tide-staflf to these bench-marks determined. This 

 would make possible correlation with any future tidal 

 observations at the same places, and might even permit 

 a quantitative determination of the local rate of eleva- 

 tion or subsidence of the land relative to the sea. 



A fertile and almost virgin field is offered to the 

 investigators of a new Challenger expedition in the 

 study of the tides of the open sea, the importance of 

 which is obvious. Several forms of automatic tide 

 gauges adapted for oflfshore tidal observations appear 

 to have given satisfactory service. Recently an oflf- 

 shore hydrographic party secured an excellent series of 

 tidal observations by means of an improvised tide 

 gauge consisting of a graduated tide-staflf secured to 

 a float and confined in a float-tube made up of sec- 

 tions of 2-in. pii>e, the lowest section of which was 

 cast into a block of cement. It also appears that satis- 



