6 ; 
and firmness of the colony, both these columns do not yield much; isolated large polyps 
occur on' colonies of three stations; on IV some very large ones appear between the already 
very unequal polyps. — The firmness of the colony can show all possible transitions between 
rather firm, upright colonies and very slender, capilliform ones; it stands to reason that 
especially in this regard the vague and relative data of former authors fail me. 
A rather important character is the angle between the branches and the 
branch which bears them, while the relative position of the branches is a specific 
character of great importance in the former descriptions. The following angles occur: 
30°: XI, XV a (one of 90°!) XX, XXI / 3 . 
30°—- 45 °: II, XII (partly), XVIII, XXV, XXVIII (rarely 90°), XXIX (rarely 90°), 
XXX (rarely 90°), XXXVIII, Antipathes foeniculacea Pall. 
30°—90° : XXIV. 
45 0 : I, V (one of 90°), XII (partly), XIII, XXI ( 3 , XXXII (one of 90°), XXXVII. 
45°—6o°: III, XVII. 
45 °— 9 o°: IX, XXII, XXXV. 
6o°: XIV (or 45 0 ), XXI «, XXXIV. 
6o°—90°: VIII, XIX, XXXIII. 
nearly 90°: X a, XXIII, XXXVII (sometimes). 
variable, but never 90°: VII, X| 3 . 
90°: IV (or 75 0 —90°), VI, XVI, XXVII (sometimes acute). 
It is obvious how extremely variable this character is; rarely one and the same specimen 
shows everywhere the same angle, and the differences between two species are usually less 
than the individual variability. It is also remarkable in how great a degree two specimens of 
the same locality can differ in this respect. 
The diameter of the axis diminishes in most cases regularly from the base towards 
the top, but some specimens show first an increasing diameter, which only after a few cm. 
diminishes. This is the case in: I, II, III, XIV a, XXI ( 3 . In a number of cases the behaviour 
of the diameter was not to be deduced out of the descriptions; in 14 cases I found an imme¬ 
diately diminishing diameter. 
The distance between the branches varies in nearly every specimen from a 
few mm. up to some cm. In a few cases the distance is always only a few mm., viz. IV (a 
few mm.), X [3 (3 mm.), XXIV (a few mm.), XXVI (4:—10 mm.) while in one case (XXVII) 
this value increases to 4—9 cm. Between this extremes there is a regular series of transitions 
to be found, so that a sharply defined limit cannot be given. 
The shape of the branches is in far and away the most cases curved or sinuous; 
in II and XXI they are parts of a circle, but this is not a great deviation ot the principal 
type. Sometimes the branches, which leave the bearing-branch at a wide angle, are so much 
curved as to become parallel or sub-parallel with the branch of lower order. 1 he entire colony 
has in this case all its branches sub-parallel with the stem (XIV, XV). It is also possible that 
the branch is curved in a direction opposite to the curve of the branch of lower order (I). A 
unilateral insertion of the branches occurs in : IV, XXVI, XXVIII (according to Brook), XXXI, 
