'ipril 17, 1884] 



NA TURE 



587 



shows markedly tlie jjeculiarity due to the bullous structure. If 

 this gray-green pulverulent matter be placed under the micro- 

 scope, it is seen to be composed of almost impalpable grains, 

 with a mean diameter of o'l mm., which are almost exclusively 

 colourless or brownish vitreous particles permeated by bubbles. 

 The bubbles are rarely globular, but often elongated, as we have 

 just pointed out, and they give a drawn-out appearance to the 

 fragments. As often happens, several bubbles are elongated 

 parallel to each other, and in this case the pore becomes a simple 

 streak ; the fragment then assumes a fibrous textui'e, which may 

 cause it to resemble at fii-st sight a striated felspar or an organic 

 remnant ; but an examination of the outline will never a'low of 

 this confusion. If we examine the temiinal contours and lines 

 of these bubble-containing fragments, we never find that they 

 are straight lines, but that they show a ragged appearance, all 

 the sinuosities being curinlinear. This mode of fracture is in 

 correspondence with the vacuolated structure, and, just as in the 

 porous pumice, the vitreous volcanic ashes are permeated by 

 vacuoles ; besides, everything goes to show that the fragmentary 

 condition and the fresh fractures are due to a tension phenome- 

 non which affects these vitreous matters in a manner analogous 

 to what is observed in the *' Rupert's drops." 



iJ, 



iMa 





Fig. I. — Vitreous particles cf the ashes of Krakatoa, which fell at Batavia, 

 August27, I883GI,,)- 



AVe have pointed out that brown vitreous fragments are rare 

 in *he ashes of Krakatoa. These, however, contain skeletons of 

 n. gnetic iron, and are devitrified by microliths. ^ It is scarcely 

 necessary to add that the particles, whose form we have indi- 

 cated, are isotropic. Iftmder crossed nicols we sometimes see 

 the field illuminated, this is due to crystals in the vitreous matter, 

 or to phenomena of tension, which are sometimes observed in the 

 neighbourhood of tlie bubbles. 



These details on the micro-structure of the vitreous particles 

 from Krakatoa can be applied with most perfect exactitude to 

 the volcanic dusts, which we have determined as such, in the 

 deep-sea deposits. In virtue of their bullous structure, their 

 dimensions, and their mode of projection, they are capable of 

 being widely transported from the point of eruption by aerial 

 currents. It must be admitted, howexer, that in the deep-sea 

 sediments a very large part of these vitreous splinters has not 

 been derived from the pulverised ejections from a volcano, but 

 from the trituration of floating pumice, of which we have given 

 above a striking example. It will be understood that it is 

 scarcely possible to trace the difference between volcanic ashes, 

 properly so called, and the products resulting from the pulverisa- 

 tion of floating pumice which we have just indicated. As in the 

 incoherent products of Krakatoa, so we find spread out on the 

 bottom of the sea many more vitreous particles, similar to those 

 we have just described, than of trae volcanic minerals. This is 

 easily explained, however, when we remember how the distribu- 

 tion of volcanic dust takes place. 



Let us now point out the minerals which can be determined 

 with certainty in the ashes of this great eruption ; and we may 

 at once remark that they are the same which we have almost 

 always found associated in the deposits with the splinters of 

 glass. In general all the ciystals are fractured, except those 

 which are still embedded in a vitreous layer ; this vitreous 

 coating is often crackled and bullous. In the ashes of Krakatoa, 



* Just as we can divide pumice microscopically according as it is acid or 

 basic, so the products nf its trituration may be recognised under the micro- 

 scope, inasmuch .ns the formrr often give colourless and more elongated 

 particles, while the fragments of basic pumice have a more pronounced tint 

 and more rounded pores. 



however, we have not remarked the globules of glass w hich are 

 often described as glued to the minerals of volcanic ashes, nor 

 have we seen the drawn-out vitreous filaments resembling Peles 

 hair. The minerals of the Krakatoa ashes which are susceptible 

 of a rigorous determination belong to jjlagioclase, augite, 

 rhombic pyroxene, and magnetite.^ We shall presently see the 

 peculiarity which distinguishes each of these species in the ashes. 



Among the most frequent minerals, but poorly represented in 

 comparison with the vitreous matter, plagioclase felspar comes 

 first. This mineral has about the same dimensions as the 

 vitreous fragments, and, with the exception of the crystals, 

 entirely inclosed in the pumice matter, is in the form of debris. 

 Sometimes twins on the albite plan can be distinguished, and 

 the results of analysis clearly indicate that it is triclinic felspar 

 which should almost exclusively be found in this ash. But the 

 most interesting crystals of plagioclase, and the most charac- 

 teristic of this ash, although represented very rarely, are in the 

 form of rhombic tables, extremely thin, and covered with a fine 

 lacework of vitreous matter. We know that the crystals de- 

 scribed by Penck " in a great number of lapilli and of volcanic 

 ashes, upon the nature of which doubts have been expressed, 

 belong incontestably to the plagiocalases, and represent an iso- 

 morphic mixture analogous to that of bytownite. It is to Mr. 

 Max Schuster " that we owe this specific detemiination. Ha\-ing 

 found in numerous sediments of the Pacific these same crystals 

 in the form of rhombic tables, and possessing preparations 

 which would be of great interest to him in his remarkable optical 

 studies on the felspars, we submitted them to this ingenious 

 mineralogist in order to confirm our determination. We believe 

 it will be interesting to give a resitvte here of the results of the 

 observations of iSIr. Schuster, which are perfectly applicable to 

 the characteristic ciystals of felspar from Krakatoa, as well as 

 to those which we have discovered in a great number of deep- 

 sea soundings. 



This plagioclase occurs for the most part in flat tabular 

 crystals with the clinopinacoid especially developed. Indi- 

 viduals of the columnar type, elongated in the direction of the 

 edge P/M, are rare. These tabular crystals consist essentially 

 of a combination of the clinopinacoid with P and x, more 

 rarely with P, ti, and y, and occasionally x and y appear to- 

 gether. In the first case the crystals have the form of a rhomb, 

 in the second case they are elongated through the predominance 

 of either x or P. The dimensions of those crystals which were 

 examined and measured lie between the line o'6i mm. broad 

 and I mm. long as maximum, and o'ois mm. broad and 

 o'042 mm. long as minimum. The extinction of the plagioclase 

 is negative. Its value was found to vary between 22° and 32° on 

 the clinopinacoid, and between 8° and 16' on the basal plane. 

 The average values of many measurements made on good crystals 

 are as follows : — 24° 12', 25° 6', and 29° 6' on the clinopincaoid, 

 10° 42' on the one side, and 10° 18' on the other side of the 

 twinning line, as this is shown on the basal plane. Polysynthetic 

 individuals, made up of repeated twins on the albite plan, were 

 very rarely observed. The felspar in its optical properties is 

 thus seen to lie between labradorite and bytownite. The twin 

 growths are particularly frequent and interesting on account of 

 the structure of the individuals. In addition to those of the 

 albite type, others were observed in which the edges P/M and 

 P/K could be definitely determined as the axes of twinning, 

 whilst P and K formed the twinning planes. The plane of com- 

 position was principally either P or M when penetration twins 

 were not observed. 



These fragments and crystals of plagioclase contain inclusions 

 of vitreous matter, and sometimes grains of magnetite. Perhaps 

 a small number of felspathic grains may belong to sanidine, the 

 presence of which is insinuated by the percentage of potass indi- 

 cated by the analysis which follow (KoO = o'97 per cent.). 



We liave said that the pyroxenic minerals of the ash are augite 

 and a rhombic pyroxene ; we distinguish them by the microscope 

 sometimes in the form of fragments — and this is usually the case 

 — sometimes in the foiTn of crystals, which we can isolate from 

 the volcanic glass covering them by treating them with hydro- 

 fluoric acid. In the crystals of augite we distinguish the faces of 



* Lately the works on these same ashes have made known as accidental 

 elements pyrites, apatite, and perhaps biotite (?). It is to be remarked, 

 however, that these minerals must be extremelyrare in comparison with the 

 vitreous matters and mineral species above-mentioned. 



- Penck. " Studien liber lockere vulkanische Auswurfiinge," Zeiischr, d 

 deutsch. ^eol. Geselisch., 1878. 



3 Schuster. " Bemerkungen ru E. Mallard's Abhandlung sur I'isomnr- 

 phisme des feldspaths tricliniques, &c.." Milt. fctr. Mitllt., v, 1882, 



