596 



NA TURE 



[August 25, 1904 



fertilised, and experiments of a similar nature carried out 

 on wild plants in situ in June, 1903, and also June, 1904. 



In no case could I get a flower to fertilise itself, though 

 crossed flowers produced abundant seed under both con- 

 ditions. A correspondent in Edinburgh, experimenting upon 

 primroses for quite another purpose, confirms my experience 

 in this matter. 



The author of the above-mentioned work has in my 

 opinion fallen into the common error of deducing a function 

 from a structure without recourse to the experimental 

 7nethod, a mode of procedure which has, I believe, led' him 

 into grave error. E. A. Bunyard. 



The Bungalow, Allington, Maidstone, August 8. 



An Optical Phenomenon. 



Mr. Hillig's letter in N.ature of .'\ugust 18 (p. 366) 

 reminds me of a somewhat similar phenomenon which I 

 observed last May when using a rotating cubical mirror 

 and sensitive flame. 



When the mirror was rotated by hand at moderate speed 

 the upper and lower edges of the band of light seen in the 

 mirror presented exactly the appearance of a faint spec- 

 trum, red being outside and pale green and blue inside. 

 The central portion of the band was colourless. 



The appearance was most distinct when the flame was 

 influenced by a sound. 



I repeated the experiment to-day with the same result. 

 George W. Walker. 



Physical Laboratory, The University, Glasgow, 

 August 19. 



Traction of Caniages. 



In further answer to your correspondent, p. 270, 

 in passing along a road the wheels of a carriage en- 

 counter many small obstacles and inequalities over which 

 they have to rise. In doing so the centre of gravity of 

 the load (which is always higher than the axles) is raised 

 to a greater vertical height when the axles are far apart 

 than when they are close together. The work done in 

 the former case is, therefore, greater than in the latter, by 

 an amount the magnitude of which is proportional to the 

 difference between the versed sines of the angles through 

 which the carriage is tilted in each case respectively. The 

 same argument applies in regard to lateral oscillations of 

 the centre of gravity with the corollary that the narrower 

 the gauge the more easily is the carriage propelled or 

 drawn. 



There may also be some question as to the influence of 

 the different rates of retardation and acceleration of the 

 centre of gravity in each case. 



Cardiff, August i. W. Galloway. 



Indian Rhynchota. 



T\ the issue of Nature of August 11 (p. 341) there 

 appeared a notice of my second volume on " Indian Rhyn- 

 chota " (Blanford series), in which I read, " the two last 

 families of the Gymnocerata (Hebridae and Hydrometridce) 

 are left over to be included with the Cryptocerata " in the 

 third volume. 



This is an error. They have already appeared in their 

 proper location, vol. ii. pp. 167 and 168-192. 



W. L. Distant. 



Steine House, .Selhurst Road, South Norwood, S.E. 



[The reviewer regrets the oversight which Mr. Distant 

 has pointed out. — Ed.] 



The Earliest Mention of Hydrodictyon. 



TwAN Ching-Shih {ob. S63), in his " Vu-yang-tsah-tsu," 

 Japanese edition, 1697, tom. xix., fol. 12, a, writes : — 



" The Shwui-mung-tsiau (literally. Water-net-alga) grew 

 in Kun-ming-chi [an artificial lake formed by the order] of 

 the Emperor Wu-ti of the Han dynasty [reigned 140-87 B.C.]. 

 Its branches, spreading sidewise, now come out of water 

 slantly. It was eight to nine feet long and so closely 

 resembling the meshes of a net that the ducks could not 

 come out of it when got therein. Hence the name." 



NO. 181 7, VOL. 70] 



This is likely to be an exaggerated Chinese account of 

 the now well known water-net {Hydrodictyon utriculatum, 

 Roth.). In this part, when a paddy-field has its water 

 drained off, we meet frequently this alga, " spreading side- 

 wise, now coming out of the remaining water slantly," 

 although such a gigantic dimension as " eight to nine feet " 

 is totally out of the question. Perhaps this is the earliest 

 record of the alga. Kumagusu Minakata. 



Mount Nachi, Kii, Japan. 



MARCONI ]VEATHER TELEGRAMS. 



METEOROLOGISTS have for a long time felt that 

 they have practically come to the limit of their 

 resources in the matter of weather forecasting, so far 

 as the weather changes in the British Isles are con- 

 cerned, e.xcept, perhaps, if finance allowed that 

 telegrams might be received at a later hour than 6 p.m. 

 This later information might avoid the possibility of a 

 storm system advancing^ towards our western coasts 

 slipping in unobserved between the present hours of 

 observation, 6 p.m. and 8 a.m., without proper intima- 

 tion of its approach by the fall of the barometer and 

 the backing of the wind being duly notified. Occa- 

 sionally some of our worst storms spread over us in 

 this way, and the forecaster, who has been unable tO' 

 foresee the incoming disturbance by aid of the 6 p.m. 

 weather telegrams, finds to his dismay when viewing 

 the next morning's weather information that the full 

 violence of the storm is upon us, for which no storm 

 warnings have been issued. In this way, from time 

 to time, the central area of an important storm is welB 

 over the L'nited Kingdom before our Weather Office 

 is aware of its existence. 



Wireless telegraphy promises to supply the missing 

 link in the connection of our shore weather system, 

 with that over 'the ocean to the westward of us, and 

 the present praiseworthy effort on the part of the Daily 

 Telegraph seems likely to prove, even now with the- 

 restricted powers of the Marconi system, that useful 

 information can be obtained. The messages at pre- 

 sent are transmitted only about 100 miles from land,, 

 but the scheme which has been most ably inaugurated 

 has provided that, in addition to the latest weather 

 report when approaching our shores, there should also 

 be a report of the weather experienced some time 

 previous, so that it is not merely an isolated observ- 

 ation with which we have to deal, but a fair knowledge 

 of the weather from about mid-.\tlantic is secured. 

 This information is at times supplied by two or three 

 vessels, so that synchronous observations are obtain- 

 able, and it will easily be understood that with further 

 development of the system an approximate synoptic 

 and synchronous map of the .Atlantic may be produced. 

 The storm areas very frequently follow a course almost 

 due east when approaching our islands, but often when 

 in fair proximity to our coasts they trend to the north- 

 east, and any help in enabling a true estimate to be 

 formed of the storm's path will be of the greatest 

 possible advantage to the forecaster. Information 

 with settled conditions will be of great value, since it 

 is expected that forecasts should show with some 

 certainty the advent or continuance of settled weather. 

 When an area of high barometer readings is situated 

 to the westward of our shores, and is willing to give 

 wav, it affords an indication of the early approach of 

 storm svstems, with disturbed weather, from the open 

 ocean, while if the anticyclone maintains its ground 

 the approaching disturbance will be fended off and 

 made to follow a course more to the north-east, and 

 mav be taken altogether beyond the limits of the- 

 United Kingdom. 



For some time past the Meteorological Office has 

 had in hand the charting of the weather over the North 



