420 Geology of 



middens of the Moa-hunters occur also there. The principal settlement 

 was however situated about sixty yards from both rivers, and judging 

 from the lines of 3Ioa ovens and kitchen middens, the dwelling places 

 followed the same direction as the terraced ground. The ovens consist 

 in the centre of five to six rows, but near the banks of both rivers they 

 diminish considerably in number. They are either empty or nothing 

 but loess, silt, or vegetable soil lies upon the stones of which they are 

 built up, or they are filled with heaps of broken bones, chips of chert 

 and knives of sandstone, this refuse sometimes also forming heaps in 

 close proximity to the ovens. All these remains are invariably covered 

 by three to six inches of soil having the character of loess. Some of 

 the ovens are of an oval shape eight feet long and five feet broad, 

 others are more circular and about eight feet in diameter. The outer 

 rim is generally built up with larger stones, similar ones fill the interior, 

 piled in four to five layers upon each other, of which, of course, many 

 by the intensity of the heat have been split into angular fragments. 

 From five to eight of these ovens are usually lying together, with 

 intervals of about 20 yards between them and the next group, the 

 ground between them having probably been the camping ground of 

 the Moa-hunters. Here large fiat stones, 10 to 12 inches long and 

 six to eight inches broad are also found, together with a longish round 

 boulder also of large dimensions ; they were doubtless used for 

 breaking the bones in order to extract the marrow. All these 

 stones without exception, had to be carried from the rivers or sea 

 shore to the plains, and their great quantity testifies that for a long 

 time, this locality must have been a favourite resort of the interest- 

 ing people inhabiting the country at that distant period. 



Proceeding to an examination of the kitchen middens or refuse 

 heaps, we observe that by far the greater portion consists of AToa 

 bones, belonging to several species, identical with those obtained in 

 the turbary deposits of G-lenmark. As pointed out in former publi- 

 cations Meionornis casuarinus is there the most common species, 

 then follow according to their number 2f. didiformis, JPalapteryx ele- 

 pliantopus, Euryapteryx gravis and rlieides, all living doubtless in droves, 

 whilst — Dinornis gracilis, strutJiioides, maximus and robustus, mostly 

 represented by a few specimens only, had probably solitary habits. 

 The same rule holds good in the Eakaia kitchen middens, where the 

 more or less frequent occurrence of bones belonging to the different 

 species enumerated above, coincides closely with observations made 

 on the subject in Glenmark. Of the smaller species of our extinct 



