154 PROF. T. W. EDGEWOKTH DAVID, MK. G. H. HALLIGAN, AND MK. A. E. FINCKH. 



and so would be much more liable to be arrested in their downward progress by such 

 a comparatively small obstacle as the horny stems of the Gorgonidcie, than would 

 cubical and rounded fragments Avhich would fall quickly to the base of the submarine 

 cliff. As the nature and mode of origin of this submarine cliff, so characteristic of 

 this as of many other atolls, appeared to us of extreme interest, we devoted more 

 time to its study than to that of any other part of the submarine slope of the atoll. 



Notwithstanding the small projections of Lithothamnion and the alcyonarian 

 " scrub " above referred to, the general surface of the reef slope, between 45 and 

 140 fathoms, must be fairly smooth, as is proved by the fact that we rej)eatedly 

 dragged our dredges over this part of the slope without their becoming foul, and we 

 were even able to drag the big iron crown with its circle of teeth (each about 4 inches 

 long) over the whole surface without its becoming entangled. This coral atoll 

 accordingly affords us the remarkable feature of a neai'ly vertical wall 5UU to GOO 

 feet in depth, largely fragmental (at all events as regards its surface), which is being 

 constantly added to, and whicli thus enlarges its periphery oceanwards, though at 

 an extremely slow rate. We, of course, realised the importance of ascertaining what 

 were the materials composing this cliff face, but as it was not practicable to moor 

 a boat outside tlie atoll with a view of attempting to bore into this slope, we had to 

 content ourselves witii breaking off pieces by percussion Avith the heavy steel chisel, 

 and recovering tliem simultaneously l)y means of tiie hemp tangles in the manner 

 described above. 



It may be mentioned that in the course of about two weeks' work in the aggregate, 

 we never once obtained a specijuen of what we might consider a piece of anaent 

 coral reef rock. We had thought it possible, for example, that we might have 

 secured fragments of some Tertiary rock like Orhitoidcs limestone or nunnnulitic 

 limestone from the deeper levels, but nothing wluitevei' of this nature was obtained. 

 All the S})ecimens collected by us were recent, mostly either a mass ot partly 

 living organisms or rolled fragments of reef rock slipped down from higher levels and 

 cemented on to the steep face by Lithotliamnion^ Polytrema, Serpulce and Polyzoa. 



At the foot of the steep slope, at a depth of from 110 to 140 fathoms, our dredges 

 frequently became foul. The material collected, whether \\ ith the sand bucket, or 

 the sand pump, or the chisel and hemp tangles, was, however, very similar to that 

 obtained from the steep cliff face. The same remark ajjplies to our collections from 

 still deeper levels, down to over 300 fathoms. As already stated a good deal of sand 

 was found at or near the 200-fathom level. 



On the whole the results of our dredging between 140 and 200 fathoms appeared to 

 supply evidence not inconsistent with the view already exj)ressed by Professor 8ollas, 

 that the convex outward curve in the contour of the atoll (as seen in section) for a 

 short distance below the base of the submarine cliff, is due to an accumulation 

 of fragmental material fallen over the cliff, and now forming a talus at its base.* 



*■ 'Nature,' vul. 55, Fel)niaiv, 1897, p. 375. 



