NATURE 



\April lo, 1890 



one open to the air, and one with air sealed in— were heated 

 together and successively to 100° C, 120°, 150°, 200°, 250°, 270°, 

 and 300°, and the zeros observed. Even then, there still would 

 remain to be explained the strange depression which I noticed 

 in several sealed thermometers of lead-glass in the neighbourhood 

 of 270°. At present, I regard the suggestion as neither proved 

 nor disproved. 



We are, in fact, only beginning to learn what silica and 

 silicates are. I have quite lately, for example, found a critical 

 |3oint in the action of heat upon fire-clays, similar to the 270' 

 point in the zeros (before referred to) of my lead-glass 

 thermometers ; and a similar point is known to exist in the 

 relation of the refractive index of quartz to temperature. 

 Results of this kind show clearly that thermometry is by no 

 means an easy subject. Indeed, I might define it as a rnixture 

 of very complicated chemistry with very complicated physics. 



Glasgow, March 28. Edmund J. Mills. 



The Shuckburgh Scale andsKater Pendulum. 



By permission of Prof. T. C. Mendenhall, Superintendent 

 of the United States Coast and Geodetic Survey, and of 

 AVeights and Measures, I enclose to you for publication, if 

 deemed suitable, a note relating to an abstract of a paper by 

 ■(General J. T. Walker, R.E., F.R.S., published in Nature 

 of February 20 (p. 381). 



As the subject-matter refers to U.S.C. and G.S. Bulletin 

 Is^o. 9, I take the liberty of enclosing it also. 



O. H. TiTTMANN. 



United States Coast and Geodetic Survey, Office of Weights 

 and Measures, Washington, D.C., March 13. 



Last summer the United States Coast and Geodetic Survey 

 published an investigation. Bulletin No. 9, on the relation of 

 the yard to the metre. 



As the re.'-ult of this investigation, values were deduced for 

 the length of certain historic standards in England which 

 differed very materially from the values previously assigned to 

 them in metric measures. 



Thus the length of the Royal Society's platinum metre, 

 certified by Arago to be i7'59/i too short, was found to be only 

 7 /u too !-hort. 



This metre was compared by Captain Kater with a certain 

 space (0-39-4 inches) on the Shuckburgh scale, and this space 

 was in turn compared with his pendulum. It is therefore of 

 interest to know whether the value deduced in the investigation 

 referred to is accurate. It is the object of this note to call 

 attention to a surprising verification of the deductions contained 

 in Bulletin No. 9. Using the equation for the platinum metre 

 found in that paier, namely — 



Platinum Metre = I m. - 7 /^ -h 9 "126 /t, / C°, 

 we find 



at I5°-98C., P.M. := i -f- 138 -8 /x; 



but at this temperature Captain Kater found the space on the 

 Shuckburgh scale 



(o~39'4i"ches) = P.M. + o"0240oinch, or o"6o96mm., 

 •whence the space in question of the Shuckburgh scale 

 = I •CXD7484 m. , and using for the coefficient expansion 

 1885 X 10"^ for 1° C, we have at i6°'67 



the space = i"ooo76i4m. 



Nature of February 20 (p. 381) publishes an abstract of a 

 paper by General J. T. Walker, K.E., F.R.S., "On the Unit 

 of Length of a Standard Scale by Sir George Shuckburgh, 

 appertaining to the Royal Society," in which he states that 

 the Shuckburgh scale was taken to Paris and compared with 

 one of the standard bars of the International Bureau of 

 Weights and Measures, by Commandant Defforges. The result 

 of this comparison reduced to 16° "67 C, and as given by 

 General Walker is 



the space = i •0007619 m, 



This agreement is perfect, more so, in fact, than the circum- 

 stances allow one to expect. 



The agreement implies the correctness of the new values 

 deduced in Bulletin No. 9 for the Ordnance metre and the 

 platinum metre of the Royal Society, and gives the value of the 

 ■metre as equal to 39 '3699 inches as therein computed from 

 Baily's and Sheepshank's comparisons, which established the 

 relation between the Imperial yard and the space on the 

 Shuckburgh scale. 



It is to be no:ed that General Walker, ignoring Baily's and 



Sheepshank's comparisons, and adhering to the Clarke value 

 39*3704+ inches, deduces the (the wriier of this thinks) erroneous 

 conclusion, that the space on the Shuckburgh scale equals 

 39*400428 inches, the value according to their comparisons 

 being 39*399896 inches. If to this value be added 0*04090 inch, 

 the amount by which the distance between the knife-edges of 

 the Kater pendulum exceeds the space 0-39^4 inches, the resulting 

 length of the Kater pendulum at 16° '67 C. is 39*44080 inches, 

 a value practically identical with that published by Kater, which 

 is 39^44085 inches. 



The Green Flash at Sunset. 



The explanation of the bluish (?) green flash of light some- 

 times seen at sunset given in your note last week (p. 495) does not 

 seem to me to be a sufficient explana' ion of all the observations. 

 If the phenomenon were due simply 10 refraction it would last 

 for only a fraction of a second, and the colour would be much 

 more blue than green. But, so far as my own observations go, 

 the colour may last for several seconds, and is a bright pea- 

 green, exactly similar to that shown by the sun many degrees 

 above the horizon in South India in Se itember 1883. To 

 produce that green, as I have shown elsewhere, all that is 

 required is the absorption due to a great thickness of vapour, 

 combined with a certain amount of dust — water dust or other. 



I siw a very pretty example of this last July when off the 

 coast of Vancouver, B.C. The air was very moist and the rain- 

 band correspondingly strong, while fine da^t was supplied by 

 the la'id breeze carrying with it particles from the burning 

 forests inland. The sky was cloudles>, but the haze was thick 

 enough to allow one to look at the sun while it was still some 

 degrees above the horizon, and the disk appeared of a brilliant 

 golden-red, gradually changing to yellow, and, finally, while 

 part was still above the horizon, it became a bright pea-green. 

 The spectrum was similar to that figured in my paper on the 

 green sun (R.S.E. Trans., xxxii. 389). 



A few days later I had a view of the sunset fmm the Selkirks, 

 where the air was very dry, the rain-band flight, l)ut the haze 

 considerable. The colours of the sun's disk were much less 

 brilliant, and never passed beyond the stage of a reddish-copper 

 tint. C. MicHiE Smith. 



73 George Street, Edinburgh, March 31. 



Foreign Substances attached to Crabs. 



I MUST of course accept Prof. Mcintosh's interpretation of 

 his own statement, and admit that he has found Molgiila arcnosa 

 frequently in the stomachs of Cod and Haddo k. This Ascidian 

 differs from the majority of its class in having allocryptic habits, 

 but I have not yet made a sufficient number of experiments to 

 be satisfied as to its edibility. It has also been a considerable 

 difficulty to me that the extensive investigations of Brook and 

 Ramsay Smith lend no support at all to the opinion that this 

 Ascidian forms an article of food for ground-feeding fish. In 

 any case the matter, though of much interest, is not one for 

 discussion here, since A/o/gu/a arenosais never one of the "foreign 

 substances attached to crabs. " 



The statement made by Mr. Holt that "Actinia mesembry- 

 anthennun is a favourite food of the Cod," was so inconsistent 

 with our knowledge of the habits and distribution of the two 

 species that, as I expected, the grounds for his assertion prove 

 to be entirely fallacious. My statement with regard to the 

 offensiveness of Actinians to fishes was made after prolonged 

 observation of the habits of the living animals and after ex- 

 periment, while Mr. Holt bases his objection on the ground that 

 the St. Andrews fishermen find A. mesetnbryanthemum to be a 

 successful bait for Cod. One might as well arguo that because 

 bits of red flannel or of tobacco-pipe are highly successful baits 

 in whiffing for Mackerel, therefore these substances form a 

 "favourite food" of this fish. A moment'^ reflection also 

 would have shown Mr. Holt thai an A-.emonc impaled upon a fish- 

 hook is a much less dangerous creature than une under natural 

 conditions and with tentacles expanded. 



During the past week an interesting ob>ervaiion of Eisig'shas 

 come under my notice which C)rroborates the view that the 

 association between Crabs and Anemones is of primary import- 

 ance for the protection of the Crabs. Eisig observed (see journ. 

 R.M.S., iii., 1883, p. 493) that an Octopus in i's attacks upon a 

 Hermit Crab would instantly retreat up <ii being touched by the 

 stinging organs of the Actinian associated with it. 



Plymouth, April 5. Walter Garstang. 



