July 1 8, 1878] 



NATURE 



301 



lines, green, blue, and indigo, was noticed also in jets of flame 

 projecting from above the lid of a coke crucible furnace in which 

 copper and brass w-ere being melted (the first of the three 

 furnace spectra in the figure), where the presence of a chloride 

 would not be so readily suspected. The same furnace was re- 

 visited, and fired for experiment in different ways, but it only 

 showed the copper-flame spectra drawn in the last two lines of 

 the figure, whose total coloiu: is olive, or tawny green, quite 

 different in aspect from the rich blue of the flame first seen and 

 identified with that occurring in ordinary fires. As refuse brass 

 and copper articles are consigned for melting to these crucibles, 

 it is, however, not improbable that copper chloride may in this 

 instance have been introduced among them. 



Attempting to discover the cause of the blue fire-flame by 

 examinations of artificial spectra led me to try that of the blue 

 flame found by Prof. Barrett (Nature, vol. v. p. 483) to be 

 produced when a burning jet of hydrogen is allowed to play upon 



Bunsen 

 flame. 



spectrum 



Spectra of the blue flame of ccpper-chloride, naturally and artificially pro- 

 duced; observed, a, with BrowTiing's miniature spectroscope with 

 reflected micrometer scale ; b, with punctured-scale pocket-spectroscope. 



the surface of sulphuric acid, or upon the surface of any object, 

 indeed, which has contracted dust by exposure to the air. It is 

 very abundant, and readily produced everywhere, and not less 

 intensely blue than the copper flame ; but no distinct measure- 

 ments or notable appearances of its spectral bands could be 

 obtained. I could, however, corroborate Prof. Barrett's obser- 

 vations of the extension of the blue colour only on the surfaces 

 touched, and its want of penetration into the body of the flame ; 

 and I noticed that metal surfaces rubbed with sulphur which 

 remain cool exhibit it more brightly than wood or other sulphur- 

 ised siurf aces which are quickly heated in the flame. A paste 

 of coke-dust and sulphur wetted with water ceases to tinge the 

 flame (as does also sulphuric acid) when it grows warm, and it 

 fails to colour it when dry. The brightest blue flame was 

 obtained by directing the burning jet upon a mixture of snow 

 with coke dust and sulphur, pounded together in a mortar, the 

 reason apparently being that great attenuation of the sulphur- 

 vapour, and therefore a low temperature of the sulphur, is 

 required to enable the sulphur compound formed to eschibit its 

 characteristic blue-coloured spectrum in the flame. 



A. S. Herschel 



Zoological Geography — Didus and Didunculus 



^ My use of the expression "a near congener" seems, from 

 Prof. Newton's letter in your number of July 4, to have diverted 

 attention from the point to which I desired to direct it. 



If the bird of the Navigator group had presented such near 

 congenerity to the Dodo as does the other of the ground birds 

 (Fezophaps) known from the Mascarine Archipelago, this would 

 have been startling, but as it is the degree of relationship (I 

 avoid the word "congener" which gave rise to Mr. Newton's 

 correction) seems to me to bear out the argument as to former 

 geographical connections which I so long ago advanced. 



Both birds belong to the Columbce and to that all but extinct 

 branch of the family to which the term "ground dove" has 

 been applied ; and the wingless condition of the Dodo has been 

 by one great living authority accounted for on the hypothesis 

 that by being confined to islands and so secured against enemies, 

 and finding food on the ground, this queer pigeon gradually lost 

 the necessity for, and with this the use of its wings, and thus 

 acquired its bulky form and ground habits. Instead of this, 

 however, I believe that both the Didida and Didunculus are 

 survivors from mezozoic times, of a great family in which the 

 characters that connect these ground birds with the winged 



Columkt were those common to a large order of wingless birds 

 that, like other orders of mezozoic life, have since perished. 

 From this ancient order of life, or from some yet more ancient 

 stem combining their common characters, the winged Columba 

 may have been evolved ; but of the order itself Didunculus has, 

 I believe, siu-vived at the eastern side (longitude 170° W.) of the 

 ancient continent to which I in my first letter alluded, and the 

 Dodo and its kindred at the western (long. 58° E.), in both 

 cases by the protection afforded by insulation. 



The application which I sought to make of this to the case of 

 the tortoises was that the presence of those reptiles in the Mas- 

 carine and Galapagos archipelagos is due to the same geographi- 

 cal change. The osteological differences between the tortoises 

 of the two regions may, perhaps, be less than those between 

 Didus and Didunculus ; but if so, this, in a lower and cold-blooded 

 grade of vertebrate life, would not weigh much ; and my conten- 

 tion is that the tortoises of the Galapagos are insulated there^by 

 survival from the eastern extremity of this ancient continent, 

 and those of Aldabra, in the Mascarine region, by survival from 

 the western, instead of from land extending across Africa and 

 the Atlantic to South America, as supposed by Dr. Giinther. 

 The fossils of the Himalayan and Mediterranean regions prove 

 that the great tortoises lived on the Europeo-Asiatic continent in 

 miocene and older pliocene times (becoming extinct dvuring the 

 latter), but this does not appear to me to negative the conclusion 

 drawn above. 



To prevent misconception, I should, perhaps, add that the 

 land tract from the submergence of which the Mascarine archi- 

 pelago and the differentiation of the immediate kindred of the 

 Dodo originated (as suggested by Prof. Newton in the memoir 

 in the Phil. Trans, of 1869, to which he refers), was in my view 

 but a fragment of the more ancient and far more extensive con- 

 tinent which (in i860) I attempted to show occupied the southern 

 hemisphere in mesozoic times ; and that such fragment, again, 

 was but a remnant of a still larger portion of this great southern 

 continent, which, as far back as the triassic period, had become 

 separated from the Australian portion, and, so late as the earlier 

 part of the tertiary, occupied much of the Indian Ocean, where, 

 during the eocene or miocene periods, it formed the cradle of the 

 human race. Searles V. Wood, Jun. 



July 13 



P.S. — Dr. Forbes speaks of Didunculus being somewhat, 

 plentiful still in Upolu. 



Smell and Hearing in Insects 



In Nature (vol. xvii. pp. 45, 62, 82, 102, 162-3), which 

 has just reached me, I see a discussion as to the senses of 

 smelling and hearing in moths, to which I add my mite as an 

 old observer. 



I do not see how any one can doubt the first. What but the 

 sense of smell directs nocturnal insects to their food ? At this 

 moment I have in my verandah a parrot, which is daily regaled 

 with a portion of a banana. Every evening I see a dozen, or 

 more, of the large Sphingidce and 'Noctuce trying to effect an 

 entrance into the cage to get at the rotting fruit, which is 

 generally invisible from the outside, being behind the flap of 

 wood that serves for a door ; the cage is only a rough box. I 

 have always found bananas the best bait to attract the night- 

 flyers, but only when they began to rot. 



Again, how about "sembling"? Here the odour must be 

 very subtle. A virgin female is instantly detected, while not one 

 " gay Lothario " will visit a captive matron. 



It is harder to say whether insects hear sounds, or feel them, 

 as the same effect would be produced on them by either faculty. 

 I have seen both moths and butterflies turn towards sound, and 

 direct their antenuDo to it, moving them to and fro. I have 

 noticed larvK — as remarked by one of your correspondents — 

 assume certain attitudes on being affected by sound ; these atti- 

 tudes are, I think you will find, generally those assumed for 

 protection or concealment ; the creatures are, in fact, alarmed 

 at the unusual — noise ? or vibration ? — which ? 



I will adduce one remarkable case in support of the smelling 

 power. Years ago I had (while residing in the North of 

 Ceylon) a lot of living Achatina panthera sent to me by the late 

 Mr. Blyth. I placed them in a breeding-cage, and, to secure 

 them from rats, suspended it to the ceiling of my drawing-room. 

 We soon noticed that every night the floor of the room was 

 covered with glow-worms, at which, never having seen them in 

 that part of the island before, and they being of unusual size 

 and brilliancy, we were much pleased. 



