616 MR. J. J. LISTER ON THE GEOLOGY 



channel is composed of a higher central area, separated by a rapid 

 descent from the lower peripheral part. It has evidently been 

 elevated in two stages. 



IV. Conclusion. 



The view of the mode of formation of the Tonga atolls suggested 

 in Darwin's ' Coral Heefs,' is that they grew during a period of sub- 

 sidence, when the volcanoes of the group were inactive, and that 

 elevation has occurred simultaneously with a recent volcanic out- 

 break *. Dana also considers that the islands were formed during 

 subsidence, and regards the elevation which has taken place as 

 merely a local phenomenon f- The additional information on the 

 group which we now possess appears to make a different view of the 

 formation of these islands more probable. It is clear that elevation 

 has been in progress for a long period, and has been renewed again 

 and again after intervals during which reefs of considerable extent 

 have grown ; it has to a greater or less extent affected all the larger 

 divisions of the group, while at Yavau and Eua reef-structures have 

 been raised to heights of over 500 and 1000 feet. 



Such an elevatory movement is quite in keeping with the pre- 

 sence of the line of volcanoes which traverses the group, and I am 

 not aware that there is an}^ evidence that these have, as a whole, 

 passed through a quiescent stage previous to their activity in recent 

 times. 



With regard to atolls in general, one reason for considering them 

 to have been formed during subsidence is the improbability of there 

 being any other basis than a subsided island on which the reefs 

 could have grown. In the Tonga Islands, however, the soundings of 

 H.M.S. 'Egeria' show that Tongatabu and JN^omuka rest on shallow 

 banks, and where the underlying basis of the islands is exposed, as 

 in the Nomuka group and at Eua, it is seen that they consist of 

 layers of volcanic materials laid out under water. The j)rocess now 

 going on at Ealcon Island illustrates the way in which a volcanic 

 mound, composed of soft material and originally standing above the 

 sea, may be reduced to a submarine bank, and so form a basis for 

 coral growth. 



In consideration of these facts it appears probable that the atoll- 

 shaped islands of the southern part of the Tonga Group have grown 

 on banks of volcanic material laid out in shallow w^ater, and that 

 there is no necessity to call in the hypothesis of subsidence to 

 account for their formation. 



EXPLANATION OF PLATE XXIII. 



Map of the Tonga Islands. The map of Eua is prepared partly from the 

 survey of H.M.S. 'Egeria' (1888). The outlines of the patches of volcanic 

 formation in the central part of the island are not accurately drawn. 



* ' Coral Eeefs,' 3rd ed. pp. 177, 186, 188. 



t * Origin of Coral Reefs and Islands,' Am. Journ. Sci. ser. 3, vol. xxx. (1885) 

 p. 99. 



