69i 



SCIENCE. 



[N. S. Vol. XXI. No. 540. 



are completely surrounded by the outer 

 encircling barrier reef, as are the Society 

 Islands just mentioned, which have several 

 ■wide passages into the lagoon through the 

 wide barrier reef. 



One is tempted to reconstruct the Gam- 

 bier archipelago of former times, and to 

 imagine it with a great central volcano, of 

 which Manga Reva and Au Kena are parts 

 of the rim which once were connected from 

 the southeast point of Manga Reva to Au 

 Kena, and thence along the line of the 

 outer islets to the northeast end of the 

 former island with a deep crater of more 

 than 34 fathoms. On the west face it was 

 flanked by smaller craters extending to the 

 western islets of the barrier reef, of which 

 the bays of Taku, Kirimiro and Rumaru, 

 and the bays of the west side of Tara Vai 

 are the eastern ridges. There were prob- 

 ably also other secondary volcanoes, of 

 which Aka Maru and the islets of the south 

 part of the lagoon are the remnants, the 

 latter all being situated on the gentle slope 

 of the southern part of the Manga Reva 

 plateau; this may have been the southern 

 slope of the principal volcano of the group, 

 on the face of which have grown up the 

 outer lines of the barrier reef and its islets. 



The existence of a great central volcano 

 would readily explain the great depth of 

 the lagoon in its different regions, as well as 

 the great depth off the outer face of ]Manga 

 Reva, depths showing slopes which are no 

 steeper nor more striking than the height 

 and slopes of the southern part of Manga 

 Reva or Tara Vai, of Aka ]\Iaru and of 

 Maka-pu, supposing them to be extended 

 into the sea. 



Mt. Mokoto and Mt. Duff drop precipit- 

 ously for more than one third their height 

 and in less than a quarter of a mile fall 

 from over 1,300 feet to the level of the sea. 

 Similar slopes are found along the vol- 

 canoes of Easter Island where there are no 

 coral reefs. The edge of the crater of 



Rana Kao drops perpendicularly a height 

 of nearly 1,000 feet in less than one eighth 

 of a mile horizontal distance ; and the east- 

 em face of the crater of Rana Roraka rises 

 vertically about 800 feet above the plain 

 of Tangariki. 



It is interesting to note how poor is the 

 flora of the Manga Reva archipelago as 

 compared with that of the more western 

 volcanic islands like the Marquesas and the 

 Society Islands and some of the western 

 elevated Paumotus. In the Gambler Archi- 

 pelago the forests are reduced to a few 

 patches extending along the small valleys 

 of the slopes of the volcanic spurs. I am 

 informed that even in the thirties of the 

 last century, when the missionaries first 

 landed at Manga Reva, the forest trees, 

 while more numerous, yet never attained 

 the luxuriance of growth that they attain 

 in the Society and Marquesas Islands. At 

 the present day, with the exception of the 

 forest patches just mentioned and a few 

 trees which have been introduced for cul- 

 tivation, the islands of the group are in 

 great part thickly covered with a species 

 of cane closely resembling that of our 

 southern states. The fauna of ]\Ianga 

 Reva is also extremely poor. There are no 

 mammals, and with the exception of a 

 'sandpiper' no indigenous birds. Sea 

 birds are few in number, and in our trip 

 in the eastern Pacific we rarely had more 

 than three or four birds aceompamdng us; 

 often only one, and frequently none was 

 visible for days. There are a few lizards 

 on the islands, apparently of the same spe- 

 cies as those in the Society Islands. 



We left Port Rikitea for Acapulco on 

 the foTirth of February to anchor off Aka 

 Maru; on the fifth we left our anchorage, 

 sounded off the east face of Manga Reva, 

 and took photographs. 



On our way north from Manga Reva to 

 Acapulco we did not begin to trawl or tow 

 until warned by the surface nets that the 



