MAGMATIC DIFFERENTIATION OF IGNEOUS ROCKS 321 
rhombic pyroxene in monoclinic.t In the same manner the well- 
known microscopic inclusions, often with idiomorphic contour, of 
titanic iron ore in hypersthene, diallage, and plagioclase of gabbros 
may be explained.? The latter inclusions were often interpreted 
by earlier investigators as older than the host-mineral, but in their 
present form they must be explained as later secretions from an 
originally solid solution. As may be understood from this account 
concerning the inclusions we must take into critical consideration 
a great number of momentums in determining the successive age 
of the minerals. 
When a substance is segregated from a solution, it often, as is well 
known, adheres to a solid already present (solid body or Fixkér per). 
The result of this is the deposition on a solid body (Fixkorper-A bsatz), 
which is also very important in the solidification of igneous rocks. 
We may here, for instance, refer to Figure 34, illustrating the depo- 
sition of spinel on pyrite; to Figure 33, illustrating the deposition 
of titanomagnetite on olivine; and to Figure 35, where in one place 
pyrite has been deposited on apatite while in another apatite has 
been deposited on pyrite. 
The individuals of a mineral, segregated from a magma at 
an early stage, frequently swam together to assemblings or aggre- 
gates, the result of which is a structure, for which I propose the 
term together-swimming structure or synneusts structure. 
This together-swimming may occur very rapidly. I refer to 
my publication “Die Sulfid: Silikatschmelzlosungen” (I, 1919), 
Figure 11, illustrating assemblages of octahedrons of magnetite in 
a bessemer-matte, consisting chiefly of Cu.S, and to Figure 28, a, 
illustrating assemblages of small individuals of zincblende in a 
slag. The solidification period of the two molten masses just 
mentioned, respectively molten sulfide and molten slag, needed 
only a very short time, at most half an hour. 
W. Wahl, ‘Die Enstatitaugite,”’ Tscherm. min. u. petrogr. Mitt., Vol. XXVI 
(1906). 
2In this connection we refer to a treatise by A. Johnsen, ‘‘Regelmissige Ein- 
lagerung von Eisenglanz in Cancrinit,” Centralbl. f. Min., Geol. und Pal., 1911, and 
by O. Andersen, “‘On Aventurine Feldspar,” Amer. Jour. Sci., Vol. XL (1915). 
3 Composed of civ, syn=together and vedo.s, neusis=swimming. 
