34 REPORT—1852, 
other. The atomic weights of silicium, cobalt, strontium, tin, arsenic and lead, are in 
the same ratio as the numbers 1:2:3:4:5:7. We do not see the necessity of this 
relation, but only the possibility. Why should fractional numbers only occur, and not 
whole numbers also? TI consider these relations only as facts; the law of the numbers 
themselves is quite unknown to us—as unknown as the absolute weights of the atoms.” 
On the Microscopic Structure of certain Basaltic and Metamorphic Rocks, 
and the Occurrence of Metallic Iron in them. By Professor T. ANDREWS, 
M.D., MRIA., ERS. 
If a thin splinter of basalt is viewed by reflected light in the field of a good micro- 
scope, it is seen to consist of a semitransparent granular mass, containing occasionally 
opake crystals of the magnetic oxide of iron and of iron pyrites. The former are easily 
recognized by their dark colour, metallic lustre, and the triangular and striated facets 
of the regular octahedron ; the latter, by their yellow colour and cubical form. The 
semitransparent portion which forms the great mass of the stone evidently consists of 
two distinct minerals; one having a resinous lustre, and in microscopic characters 
closely resembling crystallized augite; the other, colourless and with a glassy lustre, 
might be referred to certain varieties of felspar or of zeolite, These remarks apply 
to the compact varieties of basalt. 
The metamorphic rock of Portrush—an indurated clay-slate containing the charac- 
teristic fossils of the lias formation, and in external characters closely resembling 
Lydian stone—exhibits under the microscope a very different appearance. It is 
formed, in fact, of a semitransparent paste of homogeneous structure, everywhere 
thickly studded with innumerable microscopic cubes of iron pytites. These crystals 
are very perfectly formed, but so minute that twenty of them may frequently be 
counted in the space of ;35dth part of a square inch, the sides of the crystals being on 
an average not more than ;;55dth of an inch in length. Ifa portion of any of these 
rocks be reduced, in a porcelain mortar, to a tolerably finer but not impalpable powder, 
and a magnet be passed several times through the powder, magnetic particles will be 
found adhering to the magnet, in greater or less abundance, according to the nature 
of the rock. On removing these magnetic particles and placing them in the field of 
the microscope, they exhibit distinct polarity and all the other characters of the mag- 
netic oxide of iron. This mineral may be separated by the above simple process, not 
only from basalt, but from granite, clay-slate, primitive limestone, hardened chalk, 
magnesian limestone, and many metamorphic rocks. In short, it is one of the most 
widely- diffused minerals in nature, occurring in almost every rock which exhibits 
evidence of igneous action. The author was only able however to discover a doubtful 
trace in roofing-slate, serpentine and marble. 
After referring to the few instances in which metallic iron, not of meteoric origin, 
is alleged to have been observed, the author proceeded to describe the process by 
which he has succeeded in showing that native iron is by no means an uncommon 
constituent of basaltic rocks. The stone is first reduced to powder in a porcelain 
mortar, the use of metallic tools being carefully avoided in every part of the operation. 
The magnetic portions are then removed, as in the process for separating the oxide of 
iron, and placed in the field of the microscope. While in the field, they are moistened 
with an acid solution of sulphate of copper, which produces no change on the oxide, 
but immediately indicates the presence of the slightest trace of metallic iron by a 
deposition of metallic copper. On making this experiment, a deposit of copper occa- 
sionally occurred in irregular crystalline bunches, perfectly opake, and with the 
characteristic colour and lustre of that metal. With neutral solutions of the copper 
salt this deposit very rarely occurred, indicating either that the iron is covered with a 
film of oxide, or that it is analogous in properties to the meteoric alloy which precipi- 
tates copper from acid, but not from neutral solutions. If instead of the copper 
solution dilute sulphuric acid be added to the magnetic particles, a slight effervescence 
at particular points frequently indicates the presence of the metallic iron; and on add- 
ing solution of copper while the disengagement of gas continues, the latter is suddenly 
arrested, and a bright deposit of metallic copper appears at the same points. The 
jargest deposit of copper obtained was abont =1,th of an inch in diameter. The most 
‘abundant indications of metallic iron were obtained from a coarse-grained variety of 
basalt, which forms the hill of Slieve Mish in Antrim, and also occurs at the Maiden 
Rocks and other localities. Indications of its presence in the basalt of the Giant’s 
