July 23, 1874] 



NATURE 



223 



falling at the edge of the bar, ami would give rise to bright and 

 dark parts on the sensitive surf:ice ; the bright parts would be 

 extended by molecular irradiation underneath the opaque bar, 

 and would give rise to the irregular brushlike projections men- 

 tioned by Mr. Ranyard, instead of the uniform extension obtained 

 when the bar is kept a short distance from the collodion. It is 

 also possible that the irregular curved fluid surface m.iy at certain 

 points, where the bar was not in actual contact with the collodion, 

 have bent the rays of light undernealh the bar and given rise to 

 the irregular extension of the image. John Aitki;n 



Darroch, Falkirk, July 18 



I MUST confess myself at issue with Mr. Stillman as to the 

 result of his experiment with the strip of blackened wood laid 

 upon the collodion film. I have tried a similar experiment, and 

 find the images of bright objects sharply cut off. Even with a 

 film of four thicknesses of collodion and an exposure of ten 

 minutes, I cannot detect the smallest encroachment. The minute 

 brushes mentioned by me in my last week's letter only occa- 

 sionally occur, and appear to be due to a circulation in the liquid 

 film beneath the opaque object, probably caused by some chemi- 

 cal impurity, for I notice that the brushes only occur when the 

 film beneath the opaque object is soiled. 



It cannot be argued that because there is a difference in the 

 amount of irradiation in two pictures taken by different processes 

 (instruments, exposures, and other conditions being similar), 

 that therefore the spreading action must take place within the 

 film, for the plates prepared by the two processes may not be 

 equally sensitive, and the pictures may really correspond to what, 

 with the same process, would be different amounts of exposure. 

 Or again, the relative rates at which faint and intense light im- 

 print themselves in the two processes may differ. Want of 

 sensitiveness to the action of faint light is, I imagine, the reason 

 why irradiation is apparently decreased by the use of the red 

 collodion. A. Cowper Ranyard 



Vapourising Metals by Electricity 



In a paper in Nature (vol. x. p. igo) Mr. H. Hopkins gave 

 a short description of some experiments on vapourising metals by 

 electricity between two microscopic slides, and said that the 

 layer thus produced can be investigated by a microscope, and 

 employed in various ways to determine the character of the 

 metal. 



But the author did not point out 'Crie.vondcrful drawings shown 

 by the layer, chiefly when a slight gold sheet is used. 



This fact, very interesting in connection with molecular 

 vibrations, has been illustrated by Prof. Magrini in a lecture de- 

 livered at the Museimi of Florence, some years ago, and trans- 

 lated in I.a Rc~'ue Scuii/ifiipie (t. iv. p. 770), with some woodcuts 

 prepared by Prof. Magrini himself. 



A. RODIER 



Earth-s'nrinkings and Terrestrial Magnetism 



In my previous letter (vol. ix. p. 201) I gave some reasons for 

 believing that the earth is shrinking chiefly about its equatorial 

 region, and is being thrust out in the direction of the Poles, and 

 that the distribution of this force may be correlated with that of 

 terrestrial magnetism. As this view is somewhat novel and re- 

 volutionary, and if true will lead to considerable modification of 

 the theories generally held on cosmical forces, I wish to support 

 it by some other considerations. 



I must predicate, as to a great extent proved, that volcanoes 

 are not found in areas of upheaval. On this point I think the 

 evidence is conclusive, and as I have previously written about it 

 I shall not again enlarge upon it. I must predicate also that the 

 earth as a whole is shrinking. This I tried to show in my pre- 

 vious letter. It follows from these facts that the large areas we 

 know to be rising must be compensated by larger areas that are 

 sinking, and that we may in a measure map these latter areas out 

 by mapping out volcanoes ; for, ex hypotlusi, they occur either 

 in areas of depression or along the border lines of the oscillating 

 land. 



Thus occurring, and themselves with the related phenomena 

 of earthquakes, being the most vigorous proofs we have of the 

 mobility of the earth's crust, we may predicate further that they 

 will be found most actively at work where movements of the 

 earth are n'ost vigorously active, and that where they arc libe- 



rally scattered, there the earth's c rust is the most yielding. Now 

 if we examine the distribution of volcanoes from this point of 

 view we shall find that our main position is amply supported. 

 Within the Arctic circle there is only one volcano, so far jis we 

 know — that of Jan Mayen. Within; the Antarctic there is not 

 one. North of the 60th degree of north latitude we have the 

 volcanoes of Iceland, and three or four in Alaska, and these 

 only. South of the 5oth degree of south latitude we have Mount 

 Erebus and its companions in the South Shetlands, and these 

 only. Between the parallels of 40 and 60 the number of vol- 

 canoes increases considerably. In the northern hemisphere they 

 probably number over sixty ; but the vast majority of these are 

 contained in the semicircular line of volcanoes formed by the 

 Kiu-ile and Aleutian Islands, and which crown that vast area of 

 depression, the Pacific Ocean. In the southern hemisphere we 

 still have exceedingly few, perhaps not more than a dozen, 

 and these along the line of the Andes. It is in the region 

 bounded on the north and south by the 40th parallels of latitude 

 that we find volcanoes distributed in the greatest profusion, and 

 the focus of distribution is even more narrow than this, for it 

 may be bounded in fact by the 20th parallel on each side of the 

 Equator. It is here we have that region described by so many 

 writers in graphic terms, the Eastern Archipelago, with its 109 

 volcanoes in active operation. "From Papua to Sumatra, 

 every large island," says M. Reclus, "including probably the 

 almost unknown tracts of Borneo, is pierced with one or more 

 volcarric outlets. There are Timor, Flores, Sumbawa, Lornbok, 

 Bali, and Java, which last has no less than 45 volcanoes, 28 of 

 which are in a state of activity, and lastly the beautiful island of 

 Sumatra. Then to the east of Borneo, Ceram, Amboyna, 

 Golola, the volcano of Ternata, sung by Camoens, Celebes, 

 Mundanao, Mendora, and Luzon ; these form across the sea, as 

 it were, two great tracks of fire." (Reclus, "The Earth," 49S.) 

 Here also is that wonderful congeries of Pacific volcanoes de- 

 scribed by the same graphic arrthor. " The volcanoes of Abrim 

 and Tauna, in the New Hebrides, Turahoro, in the Archipelago 

 of Santa Cruz, and Semoya in the Salomon Islands, succeeding 

 one after the other, connect the knot of the Feejees to the region 

 of the Sunda Islands, where the earth is so often agitated by 

 violent shocks. This region may be considered as the great focus 

 of the lava-streams of our planet." It is within the same 

 narrow limits also that we have the most active signs of move- 

 ment in the Atlantic basin, namely, in the Little Antilles group 

 of the West India Islands. In regai'd to the two regions last 

 mentioned, there is a fact remarkably confirming the general 

 position I argued in favour of in a previous letter, namely, tliat 

 volcanoes are indicative of areas of depi'ession, and which w as 

 unknown to me when I wrote it. M. Reclus says — " It is a re- 

 markable fact thit the two volcanic groups of the Antilles and 

 the .Sunda Islands are situated exactly at the Antipodes one of 

 the other, and also in vicinity of the tioo polts of tiattt'n:nf^^ ths 

 existence of lulticli- on the snrface of the ^lohc has ban proved by the 

 recent calculations of astronomers. (Op. cit., p. 503.) 



These facts seem to me to support very strongly my conten- 

 tion that the earth is shrinking chiefly in its equatorial region. 

 Volcanoes are in my view the mediate and not the immediate 

 results of the shrinking of the earth ; earthquakes on the con- 

 trary are its immediate result. There is considerable difficulty 

 in mapping out a chart of their frequency and intensity, but we 

 may say safely that such a chart would have a deeply-coloured 

 zone in the equatorial regions, that it is there where earthquakes 

 and especially submarine earthquakes chiefly abound, and abound 

 also in their more vigorous type. This can only be if that area 

 is also the chief area of disturbance of the earth's crust. Another 

 fact which points in the same direction is that discussed by 

 Bischof, namely, that the soundings in the greater oceans increase 

 as we near the equator, this increase taking place relatively to 

 the land masses and not being merely due to the bulging out cf 

 the water in those parts by the force of attraction. So that if 

 we accept the level of Africa or the Pampas of Brazil as a 

 mean we shall find the greatest pits and hollows in the crust 

 in the equatorial region. 



In regard to the connection of this earth-shrinking with terres- 

 trial magnetism, I wish to quote one or two paragraphs from 

 Dr. Zollner's paper in the " Philosophical Magazine " on the 

 origin of the earth's magnetism, to the conclusions of wh'ch, 

 however, I cannot in any way assent. I quote him on the sub- 

 ject of the correlation of earthquakes with magnetic disturbarces. 

 He is quoting from Mr'. Lament's work. 



" Ivreil has given many cases," he says," where magnet'C dif- 

 turbances coincided wilh eaithquakes; hence he thinks — co!.. 



