NA TURE 



[May i, 1884 



thrown out 'of their fastenings, and fell shattered to the 

 ground with their chimneys and globes. 



But not only at this distance was the air vibration 

 perceptible. At Batoe-Radja in Palembang (250 kilometres 

 from Krakatoa) rents appeared in the pradjoerit's barracks 

 at three o'clock in the night ; even at Palembang, 350 

 kilometres from Krakatoa, several Government buildings 

 had to be immediately vacated as a crash was feared ; and 

 even in the Alkmaar country in Pasoeroean, 830 kilo- 

 metres from Krakatoa, the walls were rent in the house of 

 the administrator and the machinist. All this was caused 

 by air vibrations, not by earthquakes, for it is a remark- 

 able fact that these have nowhere been observed with 

 certainty in this eruption. 



Finally in the most violent explosions, air waves of an 

 astonishing length were formed. As the Meteorological 

 Institute at Batavia no longer possesses a self-registering 

 barometer, those waves would have passed unobserved at 

 Batavia, had they not fortunately been recorded by the 

 indicator of the gas-works. This apparatus is self- 

 registering, and continually marks on a paper, wound 

 round a turning cylinder, the pressure of the gas. As the 

 large gasometer was set in motion on August 26 and 27 

 by the pressure of the air waves, these oscillations were 

 marked by the indicator, and the line of pressure shows 

 that day, not the normal curves, but a number of sharp 

 points. As hour lines are marked on the paper, the time 

 at which these oscillations occured can be accurately 

 fixed, and if the time be deducted which the wave requires 

 to travel from Krakatoa to Batavia, the moment is also 

 known when the wave was formed, and the explosion took 

 place (omitting a correction for the time which elapses 

 between the moment of pressure on the gasometer and the 

 moment this is recorded by the indicator, a lapse of time 

 which is unfortunately not exactly known). The barometer 

 experiments in Europe and America show that those large 

 air waves possess almost as great a velocity as sound, 

 from which it follows that they require seven minutes to 

 travel the distance between Krakatoa and Batavia. 



I have concluded from this that the most violent 

 explosions took place at the following hours .-—August 27, 

 5h. 35m., 6h. 50m., 1 oh. 5111., and roh. 55m., Batavia time. 

 By far the most violent of these four was the explosion 

 of ioh. 5m. Then also an air wave was propelled from 

 Krakatoa which spread in a circular form round this point 

 as pole along the surface of the earth, and travelled no 

 less than three and a quarter times round the whole 

 circumference of the earth. The velocity was, as already 

 observed, about the same as that of sound, although 

 these were waves of a gigantic length (the length of the 

 lowest sound waves being about 20 metres, that of the 

 Krakatoa air wave more than a million metres). 



The eruptions, which took place at first above the sea, 

 probably became submarine about ten o'clock on August 

 27. Before that time only more or less damp ashes were 

 ejected, after that also a large quantity of mud or mire, 

 being volanic sand mixed with sea-water. The giving 

 way of the northern part of the mountain must have 

 preceded these submarine eruptions, as appeared from 

 the time at which the large tide wave, which probably 

 originated through this subsidence, overflowed the Vlakkcii 

 Hoek. This catastrophe caused a great change in the group 

 of islands of Krakatoa. To the north-west of Krakatoa 

 lies Verlaten Island, to the north-east Long Island, and 

 west of the latter lies the Poolsche Hoedje. This 

 small island has disappeared, the two others still exist, 

 and arc even larger than before, through the ejected sub- 

 stances which have settled on and around the island ; but 

 the greatest change has been undergone by Krakatoa 

 itself. The whole northern part with the craters Per- 

 boewatan and Danan, and half of the peak have sunk in 

 the depths. There only remains the southern part of the 

 peak, which has been cut in two from the very top, and 

 forms on the north side a magnificent precipitous cliff 



more than Soo metres high. Through the downfall there- 

 fore a volcano rupture has been formed which is probably 

 unique in the world. A coloured drawing of this remark- 

 able rock will be added to the large report. 



The size of Krakatoa was formerly 33=. square kilo- 

 metres ; of that 23 square kilometres have subsided, and 

 ioh square kilometres remain extant. But on the south 

 and south-west side the island has been increased by a 

 large ring of volcanic products, so that the size of New 

 Krakatoa is now, according to our survey, 15 £ square 

 kilometres. The size of Long Island was formerly 29 

 and is now 3'2 square kilometres. Verlaten Island has 

 become much larger ; it was formerly 37 and is now 

 1 rS square kilometres in size. Of the Poolsche Hoedje 

 nothing remains. 



In the place where the fallen part of Krakatoa once 

 stood there is now everywhere deep sea, generally 200, 

 in some places even more than 300 metres deep. It is 

 remarkable that in the midst of this deep sea a rock 

 has remained which rises about 5 metres above its sur- 

 face. Close to this rock, which is certainly not larger 

 than 10 metres square, the sea is more than 200 metres 

 deep. It is like a gigantic club, which Krakatoa lifts de- 

 fiantly out of the sea. 



The volcanic products of the preceding year consist 

 almost exclusively, as we have already said, of pumice- 

 stone ; only here and there among the pale gray material 

 a solitary piece of darker coloured steatite or a vitreous 

 piece of obsidian appears. Although the stone masses 

 in the crater were doubtless liquid, a stream of lava could 

 nowhere pour down, because everything was shot out of 

 the crater in larger and smaller pieces, and generally in 

 powder. 



The chemical composition of the ejected substances is 

 not yet sufficiently known, but from the analyses that have 

 hitherto been made it would appear that all the substances 

 do not contain the same quantity of silica ; probably that 

 the large pieces floating on the top of the molten mass 

 were somewhat more acid than the lava that lay deeper 

 in the crater and was ejected as powder. The ashes col- 

 lected by myself at Buitenzorg contained, according to the 

 analysis made at Batavia, 60 per cent ; a piece of pumice- 

 stone, collected on the Island Calmeyer, 68 per cent. ; a 

 small piece of obsidian from Krakatoa over 6S per cent. ; 

 and fine yellow ashes from the east coast of Krakatoa 

 even 70 per cent, of silica. There was found moreover 

 alumina 14 to 16 per cent., protoxide of iron 6 per cent., 

 chalk 4 per cent., soda 4 to 6 per cent., and a little 

 magnesia. 



In the microscopic examination of the ashes collected 

 at Buitenzorg there was found — (1) glass in innumerable 

 irregular fractured particles, generally completely per- 

 meated with vacuoles round or oval ; in some particles 

 the glass threads are bent. Those glass particles, micro- 

 scopic pieces of pumice-stone, are always present in large 

 abundance. (2) felspar, very fresh and clear, some- 

 times with distinct polysynthetic twin lines, generally, 

 however, in single crystals ; all seems to be plagioclase, 

 as the analysis shows no potash. As inclusions in the 

 felspar are found glass, apatite, augite, and magnetite. 

 (3) pyroxene, partly green and then extinguishing 

 obliquely, therefore monoclinic augite, partly coloured 

 brown, and then, as it appears, extinguishing in parallel 

 lines ; it is not yet quite certain whether these last brown 

 augites, which are present in much larger quantities than 

 the green ones, belong to a rhombic pyroxene (bronzite 

 or hypersthene), or are brown monoclinic augites which 

 lie in the preparations on the orthopinacoid. This would 

 be fortuitous; but if the orthopinacoid is much more 

 developed in those crystals than the klinopinacoid, it is 

 not surprising. I also believe I observed in one brown 

 crystal oblique extinction, while sometimes transitions 

 from brown to green tints occur. Inclusions are glass, 

 apatite, magnetite. (4) magnetite in grains and oc- 



