OF THE TONOA ISLANDS. 597 



Some miles to the westward of !>rango is the small island of 

 Nomiika-iki ( = little Nomuka), to the south of Nomuka, and 

 separated from it by a narrow channel. This island is about 

 60 feet hig:h, and one mile in its lon^^est diameter; it also is 

 formed in part of a hill of volcanic tuff, tlu^ remainder consisting of 

 atiat of calcareous sand. The la3'ers are horizontal and are made up 

 of finer and coarser beds of brown and grey ashes alternating with 

 one another. Here there are no fragments of coral mixed with 

 the ashes. A fossil univalve of tho genus Pyrula occurred in 

 these beds, and some of the finer ones contained vertical burrows (of 

 an annelid ?) which were filled in with the coarser material of the 

 overlying layer. 



As at Mango, there are no raised coral-rocks overlying the 

 volcanic beds, although broad fringing reefs surround the shores. 



The beds have evidently been laid out by water, but under 

 quieter conditions than those of the neighbouring island, and the 

 mound has since been elevated to its present height. 



A little to the south of Mango are the two islands Tonumeia and 

 Kelefasia. Capt. Oldham visited Tonumeia when engaged in the 

 survey of the group in 1890. He found it to consist of volcanic 

 tuffs forming a cliff 80 feet high, and dipping at an angle of 3° to 

 the south *. Mr. Harker finds that a specimen of these beds 

 consists of " fine volcanic ash compacted by a calcareous and ferru- 

 ginous cement into a yellow-brown rock." 



Capt. Oldham also obtained some black nodules two inches in 

 diameter, with portions of a calcareous matrix still adherent, and 

 which had presumably weathered out of the layers of tuff. Mr. 

 Harker describes these as consisting of oxide of manganese having 

 *' the general characters of psilomelane, but soft and of a low 

 specific gravity, and therefore perhaps altered." 



The presence of these nodules, supposing them to have occurred 

 naturally in the beds of which Tonumeia is composed, is very re- 

 markable. They are generally supposed to be formed only in deep 

 water ; but the situation of the island on the same shallow plateau 

 as Mango and Xomuka-iki, and the general correspondence of its 

 formation with theirs, point strongly to the conclusion that these 

 islands were all formed under similar conditions. The presence 

 of coral and other organic remains amongst the volcanic constituents 

 of these islands shows that they were formed in shallow water. 



Judging by its appearance as seen from the sea, I have little 

 doubt that Kelefasia belongs to the same class of islands. 



The island Tonua, to the north-east of Mango, w^as also examined 

 by Capt. Oldham. It is composed of yellowish-white rocks which 

 form a cliff twenty feet high, and consist of volcanic ashes with 

 little calcareous matrix, though shells of pteropods and the cast of 

 a gasteropod (Murea:'?) were found among the volcanic fragments t. 



In the central part of the Hapai group there are two islands, 



* The substance of Capt. Oldliam's report on this island and Tonua is 

 given in Mr. Harker's paper in the Geol. Mag. for June, lo'Jl. 

 t Harker, op. cii. 



2x2 



