Sept. 4, 1884] 



NATURE 



435 



4. The shelter of the observer can then be quite detached 

 from and independent of the telescope. 



In the description of the floating polar axis referred to above, 

 I mention its suitability for that disposition where one large and 

 one small mirror is used. 



My reason for not then mentioning the second disposi- 

 tion (that in which two large plane mirrors are used, one 

 of these being perforated) to which this polar axis is most 

 particularly adapted, was fear of stepping too far at once, as, 

 apart from the additional difficulty of making a large plane 

 perforated mirror, there seemed to me to be an element 

 of risk and uncertainty in its use. After reading this article 

 by M. Lcewy, I think there may not be much in my objec- 

 tions, but I cannot quite satisfy myself. In dealing with the 

 support of the large plane mirror in the disposition that I allude 

 to in the description of the polar axis, I contemplated such an 

 arrangement for it as I have in use with my three-foot mirrors, 

 this answering, as far as I have been able to see, perfectly, in 

 eliminating flexure, and as the back of this large plane mirror 

 would never wholly leave its supports, there would not be any 

 fear of flexure here ; the other mirrors offer no difficulty as they 

 are practically as in an ordinary Newtonian telescope. The 

 support of the large plane perforated mirror, when used for any 

 latitude higher than 45 , is not easily obtained ; it must rest on a 

 rim touching the face all round, hence unless it could be hung up 

 from the back in some way it might bend down and spoil the 

 image. 



It may be that the slight angle it would make would not bring 

 in flexure of an injurious kind, and this could be determined 

 by actual experiment beforehand, but at present it is an open 

 question. 



If it is not a difficulty, then I should agree with M. Lcewy 

 that this is the best disposition for many reasons — it dispenses with 

 the supports of the small mirror that cause diffraction rays, not 

 objectionable to the observer except on bright objects, but very 

 much so in a photograph, where they can impress themselves 

 from an eighth or ninth magnitude star with any exposure that 

 would be used for a nebula ; it gives better support to the concave 

 mirror : difficulties in connection with the reversal of the in- 

 strument do not come in with such force, and, most important of 

 all, with such a polar axis as I have described, the focal plane 

 might be kept very close behind the large perforated mirror, 

 giving advantages of the greatest importance from many points 

 of view. 



The covering of the mirrors by a plate of glass has already 

 been suggested and tried, but in a way that determined nothing. 

 It is a capital thing to do ; only experiment could really decide. 

 Certainly flexure could be got over by air pressure, and it would 

 be worth any trouble to get it, if not injurious to the image. 



Ealing, August A. Ainslie Common 



Earthquakes in Japan 



In the one hundred and seventy-first volume of the " Konrui 

 Shinko-Kushi," one of the oldest and finest works on Ancient 

 Japan, there are tables giving the number, intensity, and remark- 

 able characteristics of all the earthquakes which occurred in 

 Japan between the years 416 and 8S6 a.d. Unfortunately, the 

 few extant copies of this most important compilation are all 

 more or less in a fragmentary condition. It is, however, evident 

 from the context that the author intended to, or actually did, 

 enumerate many more of these natural phenomena, and it is 

 highly probable that many of his original notes have been lost 

 with the rest ; but even as it stands the work is of undoubted 

 importance, now that the Seismological Society of Japan has 

 been doing all in its power to bring forth the ancient records 

 which refer to the great earthquakes of the past. As every one 

 knows, Japan is the very hearth of earthquakes ; in 1S54 more 

 than 60,000 people lost their lives in consequence of one of these 

 great terrestrial catastrophes, and it has been calculated that 

 from ten to twelve earthquakes, each lasting several seconds, 

 occur every year, besides numerous others of too slight a nature 

 to be worthy of remark. 



The earthquakes mentioned in the work under consideration 

 begin with that which took place in the fifth year after the coro- 

 nation of Inkio Tenno (A.D. 416), and end with the one in the 

 fifth year of Koko Tenno (a.d. 886). Earthquakes occurred 

 during this period of 470 years on 640 days, but that by no 

 means gives the probable total. It seems that those which are 

 noted on the 640 days were all of sufficient importance to deserve 

 particular mention. The great care taken by the compiler in 



his tables is evident from the fact that the exact date and time 

 of each earthquake is given. Kiyoto was then the capital of 

 Japan, and most of the earthquakes mentioned took place in the 

 then Imperial City, 626 out of the total 640. Those not felt in 

 Kiyoto are spoken of only when unusually intense, in which case 

 the exact locality and amount of damage caused are given. Quite 

 recently the vernacular Japanese press, in consequence of some 

 lately published returns bearing on the subject, has devoted con- 

 siderable attention to investigating the annals of the " Konrui- 

 Shinko-Kushi," in hopes of being able to ascertain if earthquakes 

 of certain intensity recur at certain periods, in fact, they have 

 attempted to prove that earthquakes run in well-defined cycles. 

 This is by no means a novel nor even very modern idea. Wernich, 

 in his " Geographische-medicinische Studien," says that severe 

 earthquakes occur in Japan every twenty years. In a footnote he 

 adds : — " I am unable to adduce any natural or physical proofs in 

 behalf of this hypothesis. And yet the Japanese earthquakes can 

 be very readily explained by the theory of ' periodical pheno- 

 mena.' They are commonest at the times of the highest tides, 

 and in the months of January, April, and October." 



Whatever may be the truth of the suppositions and theories, 

 the Japanese journals, both the scientific and the dailies, have 

 gone to work by accepting the periodicity of these phenomena. 

 Taking ten years as the divisor, they divide the time between 

 A.D. 628 (when the records begin to be more trustworthy) and 

 a.d. 886 into twenty-six periods. The following table is the 

 result : — 



Periods Ear d '^ ake Periods Ear d '!^ ke Periods ^J^ 6 



1 o 10 6 19 3 



2 3 11 5 20 56 



3 o 12 29 21 39 



4 1 13 3 22 18 



5 2 14 o 23 104 



6 15 15 11 24 S7 



7 1 16 22 25 95 



8 o 17 10 26 100 



9 3 18 9 



It is very evident from the foregoing that the records are far from 

 being as exact as could be desired with regard to the earlier 

 centuries, or else that the physical condition of the country in 

 886 was totally different from that of 628 a.d. But to return to 

 the table, it will be seen that the intervals between the periods 

 in which earthquakes were most frequent are as follows : — 40 

 years between the 2nd and 6th periods, 60 years between the 6th 

 and 12th, 40 years between the 12th and 16th, 40 years between 

 the 16th and 20th, and 30 years between the 20th and 23rd. 

 Acting on the supposition that one period of unusual frequency 

 of earthquakes has been left unrecorded, the average length of 

 the intervals is estimated at 35 years. Following the author's 

 explanatory notes, a still more correct table can be deduced, by 

 means of which the cycle of earthquake intensity is finally put 

 at 33 '3 years. A further deduction is made that earthquakes of 

 a disastrous nature occur once every 59 years, so the next great 

 catastrophe may be expected in 1913. 



As the notes of the compiler give the date of each earthquake 

 between the above-mentioned years, it appears that earthquakes 

 used formerly to be most frequent in August, most severe in 

 May and November, and followed or preceded by violent hurri- 

 canes, electric storms, and the like in January ; 55 per cent, of 

 all Japanese earthquakes occurring during the warm season. 



Yokohama F. Warrington Eastlake 



" Udschimya sericaria," Rond., a Fly Parasitic on the 

 Silkworm 



I have been engaged during the past year in tracing the life- 

 history of Udschimya sericaria, Rond., and have succeeded in 

 making it out completely. I send you a short account of it, 

 hoping that it may not be entirely uninteresting to your readers. 

 As you are no doubt aware, in Japan and China the maggot of 

 this fly does great damage every year to the larva? and pupa? of 

 the silkworm, sometimes 80 per cent, of the caterpillars and 

 pupaa being killed. The knowledge of its life-history would 

 therefore be of great economic interest as furnishing the scientific 

 basis for guarding against this parasite. Strange as it may seem, 

 no one has, however, until recently, made any systematic obser- 

 vations on the matter. 



In 1874 m y father, Mr. N. Sasaki, who was the first to study 

 this insect, found its larva in the main trunk of the trachea of the 

 silkworm, just inside the stigma, and finally concluded that the 



