476 



NATURE 



[Sept. 30, 1875 



are emitted in excess ; or if the heat is sufficiently intense to 

 produce them largely, as in the melted metal, where the thin 

 films of oxide on its surface glow with perfect whiteness, the 

 metal itself must shine with bluish, or it may be with geeenish- 

 blue light, if the heat is only high enough to m>ake the excess of 

 green rays very strongly visible. If this should be, as I suppose, 

 the real explanation of the very curious appearance of depth of a 

 certain tint of colour, contrasting strongly in some parts of the 

 melted stream by its greenish hue with the surrounding redder 

 lights, according as the natural tinted appearance of the vivid 

 metal is effaced or diluted by the floating films of white-hot 

 oxides in lines and parts of the stream depending on the surface- 

 flow, and suggesting in some degree the idea of a transparent 

 cascade, and even from its colour of a waterfall, the process 

 often repeated in large foundries of ninning gun-metal into 

 large 'castings presents an instance of well-defined action of 

 the law of exchanges which must be constantly witnessed 

 and noted inquiringly by daily observers, and which certainly 

 presents, if a different and more natural explanation can be given 

 of its origin, to eyes unaccustomed and unprepared to receive 

 it, a somewhat surprising and otherwise unaccountable appear- 

 ance. In gun-metal, when the proportion of zinc introduced is 

 very small, the coating of the melted surface by copper oxide is 

 comparatively slow, and in melted brass it might not be possible, 

 from the rapid oxidation of zinc upon the surface, successfully to 

 observe the same phenomenon. In order to render melted copper 

 fluid enough for casting, a small proportion of alloy sufficient to 

 give it almost the colour of brass is required to be mixed with it, 

 and large pourings of the pure metal cannot commonly be made ; 

 but perhaps in small castings of this metal, and probably also in 

 those of gold, opportunities would present themselves similar to 

 that which I have here attempted to describe, of verifying the same 

 general law of radiation connecting together the qualities of lumi- 

 nosity and absorption in the surfaces of highly coloured metals. 

 Newcastle-on-Tyne, Sept. 20 A, S. Heeschel 



Changes of Level in the Island of Savaii 



While feeling some diffidence about setting myself in oppo- 

 sition to so careful an observer as the Rev. S. J. Whitmee (Nature 

 vol. xii., p. 291), I cannot allow his statements in regard to 

 changes of level in the island of Savaii, Samoan group, to pass 

 altogether unchallenged. In the month of June 1874 I spent 

 some weeks on the island, during which time I travelled around 

 nearly the whole of it on foot. Though not a scientific observer, 

 I was on the look-out for indications of change of level along 

 the coast, and it is my decided opinion that such indications are 

 quite as little apparent in Savaii as in Upolu. Mr. Whitmee, 

 whom I had the pleasure of meeting on the island, directed my 

 attention to what he believed to be a line of upheaved cliffs a 

 couple of hundred yards back from the sea, near Tufu, on the 

 south side of the island. On examining the place, after 

 parting from Mr. Whitmee, I particularly observed that the 

 floor of volcanic rock at the base of the cliffs bore exactly the 

 appearance of lava that had cooled in the open air. The creases 

 and ripples left on the surface of the lava in cooling were dis- 

 tinctly visible, which could not have been the case if the rock had 

 ever been exposed to the action of the waves. No doubt was 

 left on my mind that the floor of volcanic rock between the base 

 of the cliffs and the sea was at one time on a level with the top 

 of the cliffs, and that it had broken away and sunk several feet, 

 from some cause which I do not attempt to explain. 



I brought away the impression that Savaii was at one time 

 much ;more fully supplied with barrier reefs than at present, and 

 that recent lava-flows had extended the island out beyond the 

 reef. So far as my observations extended, where reefs do exist 

 they are terminated by points or capes of volcanic rock, looking 

 as if the lava had overflowed and cut off the reef. 



Orie circumstance almost, if not quite, fatal to the theory that 

 Savaii has been upheaved in whole or in part in recent times, is 

 that nowhere are there any signs of coral in situ above the sea- 

 level. In this respect it is very different from the island of 

 Rarotonga, in the Hervey group, which has'most unquestionably 

 been upheaved several feet, at least on the south side. There 

 the barrier reef is altogether out of water, and what was once 

 the enclosed lagoon is in some places dry land. 



In regard to the absence of barrier reefs in front of lava-flows, 

 I venture to suggest that it is more likely to be caused by the 

 depth of the water or by the recency of the lava-flow than by 

 any effect of existing submarine volcanic action on the coral insect. 

 ^ Sai) Francisco, Sept. 7 Richard Webb 



Origin of the Numerals 

 Having never met with any explanation of the origin of the 

 numerals, or rather of the figures symbolising them, perhaps I 

 am right in supposing that nothing satisfactory is known of it. 

 In that case the following may be interesting to your readers. 

 The first column contains the original figures, each containing as 





I 



many lines as the number which it is intended to represent. The 

 other columns show the transitions likely to result from quick 

 writing. W. Donisthorpe 



17, Porchester Terrace, W. 



Pugnacity of Rabbits and Hares 



I have occasion just now to keep over thirty Himalayan 

 rabbits in an outhouse. A short time ago it was observed that 

 some of these rabbits had been attacked and slightly bitten by 

 rats. Next day the person who feeds the rabb-ts observed, upon 

 entering the outhouse, that nearly all the inmates were congre- 

 gated in one corner, and upon going to ascertain the cause, found 

 one rat dead and another so much injured that it could scarcely 

 run. Both rats were of an unusually large size, and their bodies 

 were much mangled by the rabbits' teeth. 



I never before knew that domestic rabbits would fight with 

 ?.ny carnivorous antagonist. That wild rabbits never do so I 

 infer from having several times seen ferrets turn out, from the 

 most crowded burrow in a warren, ^young stoats and weasels riot 

 more than four inches long. 



It is evident that the show-fight instinct cannot have been 

 developed in Himalayan rabbits by means of natural selection, 

 but it is no less evident that if it ever arose in wild rabbits it 

 would be preserved and intensified by such means. And in this 

 connection I should like to ask any of your readers who may 

 be able to supply information upon the point, whether there is 

 any difference between the hares of Great Britain and those of 

 the Continent with regard to pugnacity. I have been assured 

 by Germans that in their country a hare will fight a good-sized 

 dog rather than run, and that it is dangerous to handle a 

 wounded individual. I do not know, however, whether or not 

 to trust these statements, and as there appear to be very few 

 examples of local varieties of instincts, it is desirable that anyone 

 who can should either confirm or deny this curious instance. 



Dunskaith, Ross-shire George J. Romanes 



Ol/R ASTRONOMICAL COLUMN 

 " 35 Camelopardi," B.A.C. 1924.— The principal com- 

 ponent of this double star is not included either amongst 

 the certain or suspected variables in Professor Schon- 

 feld's last catalogue, but there would appear to be suffi- 

 cient evidence of change to justify its being placed in the 

 former class. Variability was suspected by the Baron 

 Dembowski from his own estimates of magnitude 1865- 



