( 224 ) 



(.'i'citers are lying close tugetlier, parlly even overlapping each other, 

 and having clislurbed each other's relief. As a further cause of 

 transformation erosion took [)lace on tliose sides that were favourable 

 to it. At last every thing is covered by the ejected coarse sand or 

 grit and lapilli of the youngest, now all-dominating So|)oetan s. s. ; 

 the great lines of the i-elief would of course not have lieen etfaced 

 by tiiis cause only, bul it has contributed to make discrimination 

 very difiicult : the smaller rugosities were levcik^d, the contrasts 

 between crater-bottoms and walls became less manifest, especially 

 where the latter had alreiidy lundiled down, and at last the uniform 

 grey sheet of gravel, oidy locally interrupted by spare l)lulls, renders 

 it next to ini|)0ssible to discern ou mere eye-sight the xolcanic 

 formations of \arious kinds. 



In such cases a detailed topographical surxey of the \olcanic system 

 is si ill the best means of unra\olliiig the inliMcale relations, and of 

 pr(t|iLMiv inlei'|)r("tiiig the geological obsei'vations. Though the survey 

 of jyOl, made undei- my sup(MiiiteiKlencc, chielly by Mi'. W. van 

 DKN r>os, topographer to the Miniiig-de|)artment, remained restricted 

 to the iumiodiale neighbourhood of the newly lormed solfalara, yet 

 the map borrowed from this surveying allows ot" compleling in 

 some respects the opinions of former in\ esligators. 



KooKi)i<',Ks was the first to pronoiiiice iIk^ correct view thai the 

 mountain-ridge uordi of Sopoetan s. s., the Riudengan-Toiideroekan- 

 range, forms an old craler-edge. He look ihe Sopoelaii-mass lo l»e 

 the remains of a siiigli' old \()lcano, his Riiidengaii-xolcaiio, on w liich 

 the Scm}»oe — in his siiuimary identical with Kelelondei, — and 

 the Sopoetan are secondary and consequenlly younger craters. It is 

 less clear what he underslaiids by Sempoe or Kelelondei. 



The Sauasins resume Iheir observations in this way ^1. c. p. 59) 

 thai tliev range So|)oelan s. s. as a chief cone Kelelondei (and iManim- 

 jiorok). I am ol" opinion thai by doing so, ihey allow too great 

 independence to llie monticule designated by them as Kelelondei. Bnt 

 apart from this we need no longer doubt of the exislence of a large 

 old edge, «'alkMl by them Kelelondei-somnia, of ^vhich their Sempoe 

 forms a soulhern part, whilst the part of Kookdkrs' Ivindengan- 

 Tonderoekan-wall running along Ihe brook Masem forms a northern 

 and w^estern part. 



Starting from niv own observations and the evidence afforded by 

 the survev, 1 lliinU I may conclude that both representations have 

 to be combined, and that in reality besides Manimporok, two old 

 craters exist, as indicated on our map in dolled lines. In order to 

 agree as far as possible with the former' explorers the south-western 



