24 PROFESSOR W. J. SOLLAS. 



In the hope of obtaining further information bearing on the question of a supposed 

 change of level, I commenced to level across some of the islets in different parts of 

 the reef ; and I greatly regret that the idea of doing so did not occur to me earlier, 

 for soon after the work was begun our second boring came to an end, and our stay 

 upon the island was unexpectedly shortened. This was the more unfortunate as I 

 had no previous experience in the use of the theodolite, and was only just beginning 

 to discover the pitfalls that beset a levelling survey, when the time arrived for our 

 departure ; I should have liked to have been able to repeat my observations and to 

 have reduced them all to the level of tlie tide-mark, which was engraved upon the 

 Chiu'ch shortly before we left ; it would have been of interest also to compare the tide- 

 level of the ocean with that of the lagoon, which might readily have been accom- 

 plished by levelling from one to the other, and then back again. Tlie profiles of those 

 islets which were levelled across are represented in fig. 12 (see p. 19). 



Mateika presents features of especial interest ; its lagoon mound is of unusual height, 

 and, on surmounting its steep slope of loose sand on the lagoon side, a gently-sloping 

 plain is reached which ends against a low cliff cut in horizontally stratified foramini- 

 feral sandstone. Slabs of this rock lie at the foot of the cliff, and we are irresistibly 

 reminded of a raised beach. On levelling across, the marks left on the f^andy lagoon 

 slope by the waves at high water were found at the unusual height of 9 feet 9 inches 

 above low water, and the foot of the inner cliff stands at 8 feet 4 inches, or more than 

 1 foot below the high -water marks ; tlie plain at its base may therefore represent a 

 high-water terrace, and the sand ridge on its lagoon side may be a temporary feature 

 liable to be removed at any time and again piled up in accordance with changes in the 

 direction of the wind. The consolidated state of the sandstone exposed on the cliff is 

 not so readily accounted for. I do not understand it, and forbear from speculation. 



Another and remarkable instance of an islet bounded on the lagoon side by vertical 

 cliffs of nearly horizontal consolidated beds is furnished by Amatuku (fig. 16). 

 They are represented on the chart as rising 5 feet above high-watei' springs. 



Sandrock. 



'"'-.'QpcScoo^^^ Regularly bedded conglomerate 



S§^-^lo''S'^ of coral pebbles, 4 or 5 inches 



;^^^^^2^^§ '" diameter. 



."'.'■;■;-■_;- Sandrock. 



i£:^^:^^^gS» Conglomerabe. 



Sandrock. 



Fig. 16. — Diagiammatic section through the higoon-clifl' of Amatuku. 



ill ruins, and with astonishment found the sun shining on a scene unchanged. A rare fragment of coral 

 tossed on to the strand, here and there an upturned shib of coral breccia, and that was all ; the waves on 

 the English coast might have accomplished more. 



