WHITE SINTER TERRACES OE ROTO-MAHANA. 



175 



We have seen, too, from an examination of the terrace-basins, that 

 both their individual growth and the growth of the whole terrace 

 is upwards as well as outwards. Again, the earthy wall that sur- 



j?ig. 4. — Section through the Geyser, showing its Position in the Hill. 



a a. Hill. b b. Sinter terrace. c. Geyser. 



d. Cliff behind geyser. e. Kocky islet in the outlet. 



rounds the geyser-basin, and especially that part of it which is 

 opposite to the point of exit of the water, shows signs of continually 

 falling in, as if the water were undermining it and letting it down, 

 a condition of things that cannot fail, I imagine, to be brought 

 about by the building up of the geyser-basin on the side from which 

 the water issues. Lastly, the great size of the geyser basin (for it 

 is 90 to 100 ft. in diameter) in comparison with the volume of water 

 that flows from it, seems best explained on the supposition that the 

 earthy cliff at the back (cZ), which is some 50 ft. in height, has so re- 

 sisted the progress of the geyser into the hill, as to allow the earthy 

 walls on either side to be eaten away, the result being an almost 

 circular basin of the size described. 



Against the above view may perhaps be urged the fact of the 

 existence of the little rocky islet (e) in the very exit from the geyser- 

 basin, on the supposition that it would have been undermined and 

 carried away if the geyser had ever existed at a lower elevation, 

 and had built itself up into the j>osition which it now occupies. In 

 reply to this it may be urged that the distance between the little 

 island and the neighbouring bank is some 50 ft. (see fig. 5), a space 

 amply sufficient for the existence of a geyser even of the size of this, 

 if the water could find a more ready means of exit than at pre- 

 sent exists. Besides this, the island is of a much more rocky 

 character than the ordinary banks, and would without doubt offer 

 greater resistance to disintegration than they have done. 



A few words are necessary to explain the manner in which the 

 terrace-rims have originally begun to be formed, that is if we 

 suppose the geyser has burst out on the natural slope of the hill. 

 In fig. 6, let a be the point from which the hot water flows 

 down the slopes a b. As it loses its heat during its progress a 

 deposition of silica will take place ; and if the water at first flows 

 in a narrow channel, this must in time be filled up by the de- 



