REPOET ON THE DEEP-SEA DEPOSITS. 
303 
which the edges P/M and Vjk could be definitely determined as the axes of twinning. 
The plane of composition was principally either P or M when penetration twins were not 
observed. 
Crystallites are very often observed in these vitreous lapilli although no crystal has 
as yet been developed, and these elementary crystalline forms occur under various 
aspects. In certain cases they are merely parallel streaks, traversed by others at right 
angles, thus forming groups whose dimensions do not exceed 0’05 mm. At other times 
these fibres are simply parallel, and not crossed by others. In other cases they diverge 
at the two extremities of the group ; the arborescent disposition found among the 
microliths of certain pechsteins has, however, never been observed. Finally, they may 
be disposed as little fans, or irregularly interlaced and forming .balls ; sometimes the 
vitreous matter has undergone the initial stage of globulitic devitrification. When these 
crystallites are examined with a very high magnifying power, they are observed 
to be transparent, with a brownish tint. Around crystals of olivine they generally 
assume a regular disposition, the crystallites being arranged parallelly and perpendicularly 
to the crystallographic axis of the mineral. This layer of crystallites is sometimes 
composed of five or six rows ; at other times there are but one or two (see PI. XVI. 
fig. 2). In certain of the preparations of these vitreous lapilli the products of crys- 
tallitic devitrification are so crowded together that they render the base almost entirely 
opaque. Around these groups of crystallites the brown glass is observed to be sensibly 
decolorised, as is often the case when the pigmentary matter of the base is concen- 
trated in crystals or microliths. 
Black opaque spots without metallic reflection are often present, being more or less 
mammillated and elongated (see PI. XVI. fig. 1). When these spots are along lines 
of fissure they are almost certainly more or less dendritic infiltrations of manganese 
(see PI. XVII. figs. 1-3), but when embedded in the unaltered vitreous material they 
must be regarded as segregations of the fundamental mass, and, under very high powers, 
they can often be resolved into groups of crystallites, closely packed the one against the 
other. This is proved by the examination of the periphery of these black spots, for when 
the crystallites which constitute the groups can be observed upon the borders, they appear 
to be individualised. It has just been remarked that manganese is infiltrated into the 
fissures of these basic glasses ; the dissemination of this substance is so great, indeed, 
that it is sure to be found in all the preparations of rocks dredged from points 
where these vitreous fragments occur in abundance. It is possible, without the 
aid of chemical reaction, to recognise the manganese by its microscopic characters. 
The colour of the altered vitreous parts is red or brownish yellow, and they present the 
zones of decomposition, together with embedded crystals. The parts invaded by the 
manganese are brownish, and the transparency is then almost lost, or they may become 
in places quite opaque. The infiltrations of manganese do not affect polarised light. 
