DEFORMATION OF ROCKS 473 
the shearing motion the cleavage is rotated more in the more 
plastic layers and less in the less plastic layers. Hence, zm hetero- 
geneous rocks having cleavage,in a soft layer the cleavage more 
nearly accords with bedding than it does in a hard layer. 
To summarize: We have the absolute direction of greatest 
pressure varying; as a consequence of the difference in the strength 
of the beds we have the direction of greatest pressure varving from 
place to place ; we have the material of any given area on the limbs 
of folds rotated with reference to the direction of greatest pres- 
sure; and finally, we have variable differential motion between 
the beds, which momentarily changes the direction of the area 
with reference to the pressure. However, Professor Hoskins 
shows that these all combine to produce at any given point a 
shortening of the particles in one direction and an elongation in 
one or both directions at right angles to this. Newly developed 
mineral particles have similar orientations. At any point per- 
pendicular to the final position of the shortest axes of the min- 
eral particles there will be the property of cleavage. As the 
differential movement at any given place is parallel to the dip, 
the longest axes of the mineral particles will be in the plane of 
cleavage and in the direction of dip. The mean axes of the 
mineral particles will be in the same plane but in the direction 
of the strike of the cleavage. 
If before the folding a cleavage had already developed 
parallel to the bedding by deep-seated metasomatic change or 
by flattening parallel to the bedding below the level of no lateral 
stress, as suggested in the following number of this JourNAL, 
this would modify the direction of the secondary structure, 
making it more nearly parallel to the bedding than it would 
otherwise be. 
In areas of symmetrical or gently inclined folds and in homo- 
geneous rocks the shortening is approximately in the line of the 
horizon and the cleavage therefore steeply inclined. In hetero- 
geneous rocks the differential movement between the layers, as 
has been pointed out (pp. 207-208 of my first article), is ordi- 
narily upward for a higher stratum as compared with the one 
