THE DEPTH AND MARINE DEPOSITS OF THE PACIFIC OCEAN. 37 
The tuff does not contain any organisms, either calcareous or siliceous ; 
in all the specimens manganese oxide grains are irregularly distributed 
throughout the mass, and some specimens contain large patches of the 
oxides. 
From this same station the trawl also brought up pieces of pumice, one 
to two inches in diameter; a flat piece of black volcanic glass, very much 
decomposed, but still recognizable, coated with a continuous layer of the 
oxides of manganese and iron, the decomposition proceeding apparently 
along cracks, so that with a little effort the fragment can be broken be- 
tween the fingers; rounded pieces of decomposed rock, of high specific 
gravity, are also coated with the oxides, the rock itself having turned yel- 
lowish-brown through decomposition. 
HarpEenep DeEposit. 
From Station 4693 (lat. 26° 30.1'S.; long. 105° 45.2’ W.; depth, 1142 
fathoms ; 14th December, 1904), angular blocks of consolidated Globigerina 
Ooze, more or less covered and impregnated with manganese and iron oxides, 
were brought up. These blocks (see Pl. 3, fig. 6) vary in size from a frac- 
tion of an inch to four inches in diameter. In general appearance they look 
very much like slag, the outer surface being rough, vesicular, and mammil- 
lated. Of the specimens, the poorest in manganese oxide is a nearly pure 
white agglomerate of Foraminifera with irregularly distributed patches of 
the oxides inside the fragment, whilst the outside has the slaggy surface 
alluded to. Its specific gravity is low. Another sample, also chiefly com- 
posed of Foraminifera, is more coherent, and has a higher specific gravity. 
In more advanced stages the proportion of the oxides increases and is more 
equally distributed throughout the mass, giving it a rusty brown tinge, while 
the density increases. Other samples show a similar further increase in the 
proportion of the oxides of iron and manganese, but in none is a complete 
replacement of the carbonate of calcium by the oxides to be seen; even the 
parts purest in manganese still give a marked effervescence with acid. 
As regards structure, there is no tendency towards an arrangement in 
concentric layers around a common nucleus. The very few mineral par- 
ticles present in these concretions are very small, and belong to the chlorite 
group and other products of the decomposition of ferruginous minerals. 
Though the above concretions are not actual manganese nodules, they 
mark at least one step towards the formation of nodules, and this under 
