( 404 ) 



cinder or scoria specimens sent me at the time by Dr. Schocfi there 

 were also grey, white, and yellow pieces which are exclusively taken 

 from cone 4, whilst the stream and cone 5 consist entirely of the 

 dark cinder- or scoria-material, reminding Dr. Schoch of cokes, which 

 in petrographical nature seems to agree with the material of Sopoetan 

 proper. Want of time prevented my further examination of these 

 specimens. A serious argument against such a secondary, [)urely 

 hydro-lhermical mode of origin might l)e drawn from the fact that 

 tire-i)henomena have been observed. Dr. Sciioch speaks even of a 

 mass of flames which was occasionally seen : this may however be 

 occasioned by the red-hotgrit thrown up by (he \iolenl ebidlition. 



Something of this natui-e likewise might have laken jilace on a 

 smaller scale at tlie solfatare of 1901, if it had also been submerged, 

 and the stream of blocks had not come to a stop a few paces back. 

 It seems that this solfalai-e has gradnally subsided, but in 1901, 

 according to Dr. Scnorn's Malay itinerary, it was again in activity 

 with the old strength, so ihat for liim too it was impossilde lo 

 descend into ihe hole on accouui of (he sul|»hur-\apours and Ihe 

 excessive heat of the bottom. 



With sufticient certainty we may conclude from the information 

 in hand that the "cokes"-mass was not in a \u\\\u\ slate at the tiuic 

 of ejection. I'iic ciuei-aliou has taken [ilace uudergi'oimd. liolh in 

 this respect and in the manner of eruption and moving onward, 

 which is to be conceived ralhei- as a iop[>ling over each other than 

 as a moving on by proi»elling, the ])henomenon forms a I'epetition of 

 that of Roeang iu 1904 fO. c. p. 223/224 and 25(3/258). The diuerences 

 originate in the pre-existing relief of the surface and the relative 

 ciuantity of the ejected matter: nf liOcuiKj a comparatively sn^all 

 crater which was filled and from which the mass desceuded along a 

 fold of the uiouulaiu-tlauk, Iwrc a comparatixely large valley-basiu, iu 

 which the mass rolled down from one or two ei'uption-seats and 

 extended itself freely, and was not large enough to pile itself up to 

 a somewhat important height against the walls, so that the cones 

 risen on and round these eruption-seats could retain iheir dominating 

 relief, whereas at Roeang, if originally they have likewise been foi'med 

 in the space of the crater, after the tilling up, they could no more 

 be detected. In so far as 1 can remember from the ()hoiograph sent 

 bv Dr. ScnocH, the relief form of the stopped mass is what it ought 

 to be under these circumstances, a gradually sloping heap of stones 

 culminating in the two cone-slia|)ed elevations, and towards the free 

 sides a comparatively low and not too steep talus, round which 

 still some loose material is spread. It is to be presumed that in the 



