THE DEPTH AND MARINE DEPOSITS OF THE PACIFIC OCEAN. 145 



of smaller ones grafted the one on the other. On the one side of the nodule 

 the mammillae have a relatively smooth surface, on which rubbing will de- 

 velop a shining lustre ; on the other side the mammillge are more dull, with a 

 shagreen-like appearance. 



A few remains of organisms can be seen attached to the nodules, mostly 

 in the crevices separating the mammillae, and appear to be confined to the 

 side with the dull surface, which is also the side on which the greater quantity 

 of clay is still adhering. 



All these nodules develop cracks within themselves, the cracks being ori- 

 entated along the radii of the sphere, from centre to periphery. In this re- 

 spect they resemble the septaria of sedimentary strata, but unlike them are 

 not subsequently filled by deposition from solution. These cracks were most 

 certainly developed when the nodules were on the sea-bottom, as is conclu- 

 sively proved by the fact that their sides are covered with some of the clay 

 in which the nodules were formed. 



As regards mode of formation, in no case could a nucleus or centre of ac- 

 cretion be detected, but there is a typical concretionary structure in layers 

 (reminding one of the structure of an onion). One layer of the hard, black 

 manganese-iron oxide is enveloped by a layer of loose, practically incoherent, 

 palagonitic substances containing only very little of the oxides, then comes 

 another layer made up mostly of the oxides, then again a layer poor in oxides, 

 and so on. It shows in a conclusive manner that the harder and blacker 

 layers of the more compact nodules dredged elsewhere are also the layers 

 richer in oxides. If infiltration and replacement had been allowed to go 

 further in the present case, so as to bring on the formation of hard compact 

 nodules, the two sets of layers would probably still have been recognizable. 



To this alternation of layers is due the fact that the nodules not only 

 break along the cracks already referred to, but break also in a tangential 

 direction along the layers of incoherent palagonitic material. Under the 

 microscope the latter is seen to contain, besides a yellowish palagonitic sub- 

 stance, a few crystalline particles, 0.02 to 0.05 mm. in diameter, apparently 

 augite, olivine, and felspar, together with a few microlites of a colorless min- 

 eral, birefringent and extinguishing nearly straight. Magnetite is absent, 

 and no cosmic spherules could be detected. 



The successive layers of palagonitic material appear to have, as a rule, a 

 constant composition, but in one case one of the layers is different from the 

 others, being light yellow and formed of much finer material. Dendritic 



