168 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
rence of feather-fracture is found where two axial planes exist at 
different levels in the same joint-block with reversed fractures. In 
this case the plumose lines of the proximal sides of the splitting 
figures traverse the intermediate area at au angle of about 45° to the 
adjacent axes, as shown in Plate 1, fig. 2. It follows in cases of this 
kind, if our reasoning is correct, that on one and the same joint- 
plane, the separation of the rock may begin at points separated by 
appreciable distances and the fronts of the growing fracture may 
travel towards each other and pass on different levels. These twin 
fractures need further study. I have been able to find only one joint- 
block on which they were displayed, and this sjmcimen does not 
exhibit the limits of the fracture. 
Distribution of featlier-fracture on joint-blocks : —There is no 
necessary regularity in the direction of the splitting figures displayed 
upon the joints which bound a joint-block. The joints on opposite 
sides of a block and in the same parallel set may exhibit feather- 
fractures running in the same or in different directions. The feather- 
fracture may diverge from the diagonal corners of a block and 
converge at the other diagonal corners, as illustrated in Plate 1, 
fig. 3, from a specimen in my collection. 
These variations in the run of the feather-fracture in various joint- 
blocks arise from the irregularity in the length of joint-planes, and 
in their having the centers or points of origin of the splitting irregu¬ 
larly disposed. 
Feather-fracture on intersecting planes: —Where two planes 
intersect at a very large angle, feather-fracture may be traced to the 
limits of the respective planes, as shown in Plate 1, fig. 4. 
Outline of joints: — The ordinary form of joint arises where the 
fracture is limited to a single stratum. Joints of this kind are 
greatly elongated elliptical areas of fracture, as exhibited in the 
Mystic River district. Such a joint is indicated on Plate 1, fig. 10, 
which is partly ideal, being based on a study of visible portions of a 
number of joints. The central area or point of origin of the feather- 
fracture has an ill-defined boundary. The joints which traverse more 
than one stratum appear to have a similar outline; though in those 
cases where the axis of the feather-fracture is wandering and not 
parallel with the plane of the bedding, there is a lack of agreement 
in direction between the major axis of the ellipsoid of fracture and 
the horizontal extension of the stratum. This ellipsoid apparently 
is due to the fact that the joint arises in a stratum the elastic limit 
