1918.] The N.Z. Journal of Science and Technology. 
89 
A little seepage water is seen in the gully near the lighthouse, but the 
water-supply of the two villages was evidently the spring near the landing- 
place, on the eastern side of the isle. 
A few shells of Glycymeris and Cantharidus were noted on the beach 
of the isle, though none were found in the middens. A few shells of Meso- 
desma australe, Chione stutchburyi, and C. costata seen on the beach were 
doubtless the effect of drift. 
The Moana-a-kura Midden. 
The northern end of the ridge forming Somes Island narrows consider¬ 
ably, dips somewhat steeply, and then continues as a narrow spur, the 
crest of which is fairly level for about 100 yards. Across this spur, about 
half-way along, a trench or fosse about 8 ft. wide has been dug, the spoil 
therefrom having been utilized to form a rampart on the northern side 
of the fosse. This inner scarp is now from 3 ft. to 5 ft. in height, having 
suffered much from erosion, and has been destroyed in one place by modern 
roadmakers. The sides of the spur northward of this defence have been 
formed into small terraces to serve as hut-sites. Supplementary defences 
consisting of stockades would protect the terraces and prevent passage 
along the spur. Holes are seen wherein stockade posts have decayed, and 
a much-decayed post-butt was dug out in one place. Apparently this post 
had been set in an oblique position, leaning outward over the steep western 
slope, which has led one to conclude that the defence here was that singular 
form of outward-leaning stockade sometimes utilized by the Maori, and 
termed by him a matahcio. The defended area is about fifty paces in length 
and of a varying width. Both the eastern and western sides of the spur are 
abruptly steep, and could be easily defended. Shell refuse is seen on the 
slopes below. The slope down to the bluff overhanging the beach at the 
northern extremity of the spur 'shows five terraces utilized as hut-sites, 
and there is also shell refuse. 
At the base of the cliff on the western side of this old fortified position 
is a talus midden which, when examined, had just been exposed to view by 
some excavation performed by the prisoners. The shell and other refuse 
has been thrown over the cliff-head from the village above, and found a 
resting-place on the talus slope below. There it had become covered and 
concealed from view by debris from the cliff-face, probably since the village 
was abandoned. 
The most numerous species of shells here are Mytilus canaliculus and 
M. edulis, evidently of local origin. A considerable number of the shells 
of these species are still sound and entire, a rare occurrence in this district. 
Those of the first-named species, being the stronger shell, are the more 
numerous among entire specimens. These whole shells are very much more 
numerous here than in the upper or Haowhenua midden, and this may be 
owing to their being much better protected from molestation, or to the fact 
that they betoken a later occupation. The element of exposure has to be 
considered, and the lower midden may have become covered soon after the 
abandonment of the village, which would tend to preservation, whereas the 
upper midden, situated on a very gentle slope, received no protecting deposits 
of detritus and may not have been covered until Europeans levelled the site 
for the old Quarantine Station. The relative date of middens is a difficult 
question to settle. Although we know that some of Ngati-Awa were living 
on the island in 1835, when they cut off the “ Rodney ” and compelled the 
captain to transport many of them to the Chatham Islands, yet no articles 
of European manufacture were found in the two middens examined. 
