401 
interference, they will terminate in an angle of 60°, or of half the 
proper angle of the hexagon, thus forming maccles whose plane of 
union is inclined to the principal axis of the ray at an angle 
of 30°. 
In fig. 5 we have a central hexagon, carrying symmetrically on 
each of its angles a wedge-shaped five-sided table, which thus takes 
the place of the acicular ray in fig. 1. Three of the sides of this 
table appear to be those of the regular hexagon, while the remaining 
two sides would, if continued, meet at an angle which I have not 
been able to measure, somewhere within the central hexagon. From 
each of the free angles of the wedge-shaped ray, a little hexagonal 
prism is developed. 
The simply pinnate rays of fig. 2 occasionally become doubly 
pinnate. I have found even trebly pinnate varieties. 
I also noticed a form (fig. 6) which is not exactly traceable to the 
type , and which I am inclined to view rather as an abnormal condition 
induced upon one of the other forms by the coalescence, or imperfect 
development, of the secondary arms. It consisted of a six-rayed 
star, in which the rays, instead of being prismatic or wedge-shaped, 
were lanceolate, the curved boundaries thus bringing this modifica- 
