132 



NATURE 



[September 23, 1920 



The Extnoor Earthquake of September 10. 



EARLY in the morning of Friday, September lo, 

 tliree slight carthqualces were felt in North 

 Devon in the district between Morte Bay and Exmoor 

 Forest. The first occurred at about 12.15 a.m., and 

 this was followed about an hour later by two slighter 

 shocks separated by an interval of ten minutes. 

 Though it is reported that part of a chimney was 

 dislodged at East Down, and that the light at Bull 

 Point lighthouse was extinguished, it is doubtful if 

 the intensity of the shock at any place exceeded the 

 degree 4 of the Rossi-Forel scale. 



The chief interest of the earthquake lies in its 

 connection with a similar, but slightly stronger, shock 

 which occurred on January 23, 1894 {Geol. Mag., 

 vol. iii., 1896, pp. SS3-56). This earthquake disturbed 

 an area 30 miles long, extending from Ilfracombe to 

 about 2 miles east of Dulverton, and 165 miles wide, 

 the whole area containing about 389 square miles. 

 The centre of the inner isoseismal (intensity 4) was 

 half a mrle south-west of Simonsbath. From the 

 dimensions and relative positions of the isoseismal 

 lines it was inferred that the earthquake was due to 

 a fault passing close to Simonsbath, running in the 

 direction K. 22° S., and hading to the south — a posi- 

 tion which agrees almost exactly with that of the 

 northern boundary fault of the Morte Slates. The 

 length of the fault-displacement was probably about 

 10 miles. 



From the accounts so far received it appears that 

 the disturbed area of the recent earthquake is shifted 

 to the west. It is about 22 miles long, ni miles wide, 

 and contains about 230 square miles. The centre lies 

 half a mile north of East Down and a mile south of 

 the northern boundary fault of the Morte Slates — 

 that is, on its downthrow side. The length of the 

 seismic focus was about 8 miles, with its centre 

 10 miles w-est of that of the earthquake of 1804. 

 The points of interest are (i) that, as is so often the 

 case in British earthauakes, the epicentre migrated 

 to the west, and (21 that the lengths of the two foci 

 and the distance between their centres were roughly 

 equal to that which separates the epicentres of British 

 twin earthauakes. C. Davison. 



Aids to Forecasting.^ 



ALONG-FELT want of the weather forecaster has 

 been a methodical classification of weather types 

 associated with the various weather conditions which 

 present themselves. The weather-chart commonly 

 offers a picture familiar enough to the forecaster, but 

 there is much connected with the movement and 

 development which depends upon the bounding condi- 

 tions. A low-pressure system or depression when 

 appearing off our south-west coasts may have a clear 

 oath to the north-eastward, the high-pressure system 

 in its front giving way to its progress. On the other 

 hand, the hiijh-pressure system or anticyclone may 

 be well established and may maintain its ground, thus 

 compelling the advancing disturbance to adapt its 

 track to the situation — a feature of no uncommon 

 occurrence, but one offering considerable difficulty to 

 the work of the fot;ecaster. 



Lt.-Col. Gold has made a good attempt to classify the 

 different types of weather which present themselves, 

 and it seems an advance on any previous effort in this 

 direction. The memoir is not hampered by mathe- 



^ **Typc« of Pressure P-stribution. with Notes and Tables for the 

 Fourteen Years iori5-i8." Hv E. Gol'i. \feteoroloeical Office. Geo- 

 physical Memoirs, No. 16. (Published by the Air Ministry.) Price ar. (e/. . 



NO. 2656, VOL. 106] 



matics, and is the outcome of a classification and 

 analysis made by the Meteorological Section, R.E., 

 in France during the late war. It is an efTort to 

 supply the forecaster when in doubt with reference to 

 some previous situations of a similar character, so 

 that he may see what developments occurred. 



Forecasting is admitted to be a matter of experience 

 which is not always very lengthy. Fifteen types and 

 sub-types have been selected, and are graphically 

 shown in the diagrams. The controlling feature in 

 the type is the distribution of atmospheric pressure, 

 and especially the position of the anticyclone. The 

 Daily Weather Charts for fourteen years, 1905-18, 

 have been analysed according to these types and are 

 classified by months. The results are given in tables, 

 and the forecaster, having drawn his chart, can 

 see to which tyf)e it most nearly corresponds, and 

 then look up the dates in the corresponding month 

 on which the type previously occurred. On reference 

 to the Daily VVeather Charts for these days he will 

 be able to trace the later developments. 



The type-frequency is given in a table for the 

 several months on the totals for fourteen years, which 

 shows clearly the preference of different tyf)es for the 

 separate seasons. The essential weather features 

 associated with each type are clearly set out, and 

 relate primarily to the winds and weather of North- 

 East France and Flanders, although applying also to 

 the conditions over the British Isles. C. H. 



American Work in Genetics. 



nPHE increasing intricacy of genetic problems and 

 -*■ the volume of contributions from American in- 

 vestigators are notable features of present-day biology. 

 Selection of a few recent papers will indicate the way 

 in which genetic experiment is permeating many 

 fields of biological research. 



Dr. F. B. Sumner is continuing his studies of the 

 Californian races of the deer-mouse, Peroniyscus maiii- 

 culatus, and in a paper on geographic variation and 

 Mendelian inheritance (Journ. Exptl. Zool., vol. xxx.. 

 No. 3) reaches conclusions which, if substantiated by 

 further work, will be of great interest. The sub- 

 species studied are rubidus, which occurs near the 

 coast northwards from San Francisco Bay; Gambeli, 

 a coast-form south of the bay; and sonoriensis, a 

 desert form from the interior o'f Southern California. 

 Wild mice were trapped from eight stations within 

 these areas, and caged mice are also being exten- 

 sively bred. Significant racial differences are found 

 in respect to mean length of skull, ear, foot, pelvis, 

 femur, and tail, width of the dorsal tail stripe, colour 

 of p«lage, pigmentation of feet, and number of tail 

 vertebra. Local differences also occur within the 

 range of the same sub-species. Mutations have been 

 described which show Mendelian inheritance, but in 

 hybrids between geograpliic races it is claimed that 

 the result is a blend, with very little evidence of later 

 segregation. 



.4mong the numerous genetic studies of extra bristles 

 in Drosophila, that of MacDowell has been most 

 extensive. In two of his latest papers (Journ. Exptl. 

 Zool., vol. xxiii., No. i, and vol. xxx., No. 4) he has 

 further analysed the effects of selection on the number 

 of bristles. After selection for a high number 

 through several generations, reverse selection was 

 found to be impossible, except after crossing with the 

 normal type, which has four bristles. One main 

 factor determines a monohybrid ratio in crosses with 

 normal flies, but there is no dominance in the ordinary- 

 sense, and there are additional genetic differences 

 between flies having extra bristles. Studies by Pavne 



