January 20, 1898] 



NATURE 



275 



trees absorb rather than transpire moisture from the 

 atmosphere at certain seasons ? What is the transpira- 

 tion of plants at night as compared with that during the 

 day? Is there such a thing as subsurface dew, and what 

 is its amount? What is the function of stones in arable 

 land as fertilisers ? What do we know respecting the 

 diurnal flow in rivers, and the semi-diurnal rise and fall 

 in certain wells? Will a squad of men walking up to 

 the walls of an observatory, or the load equal to the 

 weight of an average man at the base of a pier in the 

 same produce any appreciable change of level on the top 

 of such a pier ? Are high mountains measurably de- 

 flected by wind pressure? In the workings of a mine 

 beneath the sea what data do we possess respecting de- 

 flections in the roof due to the rise and fall of a super- 

 incumbent tide ? What is the natural period of vibration 

 of a chimney or building of given structure? 



These are examples of the varied questions which 

 have been placed before the seismologist, and to most of 

 which, as the result of experiment and observation, he is 

 able to give fairly definite replies. Although in a few 

 instances these replies may not have given the assistance 

 to his investigations expected, a consolation remains that 



B. -Moves e.^st from 8 a.m. to 3 or 4 p.m 



~ ' • 6 5 Noon tX 18 15 13 



Fig. io. — Irregular diurnal waves in Japan, always showing tremors from 

 9 p.m. to 6 a.m. (Milne.) 



they have been of considerable value to workers in other 

 fields. 



Perhaps the greatest trouble against which the 

 working seismologist has had to fight have been the 

 ubiquitous, so-called, "earth" tremors (see Fig. lo). 

 Sometimes the apparatus will swing, and perform for 

 hours or days various irregular, and sometimes marvel- 

 lously regular back and forth movements, with the result 

 that all traces of important phenomena have been 

 eclipsed (Fig. 1 1). Not only do " tremors" affect finely con- 

 structed horizontal pendulums, but in all probability they 

 affect magnetographs, the delicate balance of the assayer, 

 and accelerate or retard the swing of pendulums. They 

 are frequent in winter, at night, and whenever the observ- 

 atory in which they are recorded is crossed by a steep 

 barometric gradient. They are particularly noticeable 

 with a frost and a falling thermometer. With a howling 

 gale, and even during a typhoon, when buildings shake 

 and shudder, they are as likely to be absent as present 

 (Fig. 12). _ 



Because a light horizontal pendulum is more disturbed 

 than one that is heavy, and that we observe at one 

 station tremors are marked, whilst they are only shown 

 feebly, or are entirely absent at another station a few 



yards distant, it is evident that we are dealing with 

 movements due to currents of air, rather than with 

 movements due to tremors in the ground. Although we 

 fully recognise, as stated at the commencem-nt of this 

 article, that there are microseismic movements in the 

 earth, it is very doubtful whether in the thousands of 

 observations carried out day and night, especially in 

 the Italian peninsula, these movements have been differ- 

 entiated from those which are the result of atmospheric 



ShiOE 



Fig. II. — Pulsations or very regular tremors \vnth periods of 2 or 3 minutes. 

 The recording pendulum has always a natural period of about 15 

 seconds. (Milne.) 



circulation. Simply opening or closing the door of a 

 case covering an instrument will sometimes start or stop 

 a so-called tremor storm. The fact that placing a tray 

 of calcium chloride inside one of these cases will cause 

 very large and continuous movements which cease on 

 its withdrawal, indicates that air currents are partly due 

 to the manner in which an atmosphere becomes dry or 

 moist. In the search for the originating cause of tremors 



W^t>i0mmm^^^0^0mm^^^mmm0tmmmm0m 



Oct IQf.H 189 5 





'j),eyf^fA/{>^^>'t^^ 



Fig. 12. — Commencement of a tremor storm at 10 p.m. with movements 

 having a period of .several minutes. At 15 and 16 hours they are very 

 irregular, whilst later the band is blackened by these movements. 

 (Milne.) 



which has extended over a period of nearly thirty years, 

 we see an excellent example of long-continued and 

 patient work which, so far as the advancement of 

 seismology is concerned, appears at first sight to have 

 terminated in a fiasco. Now, however, we know that 

 we have to distinguish between movements caused by 

 the atmosphere and those coming from the earth. We 

 see a reason why at particular observatories magneto- 

 graphs are sometimes on the swing. The reason why, 



NO. 1473. VOL. 57] 



