158 T. Crook—Titaniferous Volcanic Rocks. 
interest and importance were, however, fully recognised by Haiiy 
and other early mineralogists; but they find no place in the 
bibliography of Fouqué & Lévy’s Winéralogie Micrographique, and 
they have not been adequately appreciated by modern writers on 
petrography. 
Since these three memoirs constitute a ‘suite,’ and as each forms 
part of a definite scheme of work by Cordier on the subject of volcanic 
rocks, it seems desirable that memoirs 1 and 2 should be summarized, 
especially as their main purport has an important bearing on the 
petrographic work dealt with in Dr. Washington’s preliminary notice. 
Memoir No. 1 
This embodies the results of the examination of a large number 
of black magnetic sands, obtained from river beds, and lake and sea 
shores, in the vicinity of volcanoes. Cordier regarded the sands as 
the disintegration products of the volcanic rocks with which they 
were associated. They were noted as being sufficiently abundant on 
certain beaches to tempt people to use them as iron-ore. They were 
rarely pure (presumably as iron-ore), being usually mixed with grains 
of many other minerals, including felspar, mica, pyroxene, amphibole, 
melanite (grenat noir), and leucite (amphigéne), while sometimes 
corundum, zircon, spinel, and sphene (titane siliceo-calcaire) were 
present. 
The highly magnetic portion of the sand was separated from the 
rest by means of a bar magnet, and it was submitted to special 
examination, including chemical analysis. The portion thus with- 
drawn was described as being strongly magnetic, some of the grains 
exhibiting polarity. Their specific gravity varied from 4°59 to 4°89, 
and in nearly all their physical characters they differed from ordinary 
magnetite (fer oxydulé), with which, up to that time, they had been 
confused. 
He showed by analysis that these black highly magnetic grains 
contained a variable but fairly high percentage of titanium, and gave 
the following analyses :— 
Niedermennich. Teneriffe. Puy. 
Oxide of iron ... 79-0 fi 79°2 aS 82-0 
Oxide of titanium bee 15-9 14°8 550 12°6 
Oxide of manganese ... 2°6 1°6 es 4°5 
Alumina a as 1:0 sae 0:8 an 0°6 
Chromic acid ... ae — ae — ois traces 
We need not here discuss the merit of these analyses, except to 
point out that, in their essential features, they may reasonably be 
regarded as being fairly correct. Perhaps the most interesting 
feature is that chromium sesquioxide was observed to occur in 
traces, and it is not unlikely that this occurred as an isomorphous 
intergrowth of chromium spinel with magnetite. 
Cordier examined, altogether, 27 samples of sand from the volcanic 
rocks of ten different countries. His localities include Central France, 
Eastern Spain, the Rhine, Rome, Vesuvius, Etna, and St. Pierre 
(Martinique). In no case did he find titanium oxide to be absent 
from the highly magnetic constituent ; the amount was variable, but 
did not exceed 16 per cent. He concluded that this mineral was 
