24 
to speak only of one single spiral, and even more impossible to make use of this spiral as a 
specific characteristic. 
Further the anatomical structure of the polyps is also made use of, but although I laid 
much weight to this factor, I used it sparingly in the specific diagnoses, from obvious 
practical reasons. 
As much as possible with the description of the characters I made use of numerical 
data, for especially in the older descriptions the lack of objective numbers is to be regretted 
because of the subjective opinions of: “rather long”, “rather thick”, “rather distant”, etc. 
Schultze, and after him Roule, also followed this safer method of stricter description. — As 
to the method of measuring, as a rule I measured the length of the spine from the middle of 
the base of the spine as far as the top of the spine, while the distance between the spines is 
given between the middle of the bases of two successive spines from the same row. The number 
of longitudinal rows is always given for one aspect of the axis, whereby I always paid attention 
whether the bases of the spines of the row were visible- the rows the spines of which were 
only visible as far as their apex, did not count, as was the case with many former authors, 
since, through the very great variability of the length of the spines, in this manner statements 
are compared which must not be compared at all ; if all other data were the same, one 
species with long spines would have more longitudinal rows, visible from one aspect, than a 
short-spined species. 
As type of a spine I took the spines on those parts of the colony where the spines 
have their maximal development, since, as Brook (i) remarks, the spines at the top of the colony 
are not fully developed and in the basal part of the colony are often already diminished in 
length, since the axis increases with thicker layers as the spines themselves. 
The interpolypar distance is measured between the middle of the oral cones of the 
most neighbourly polyps if they are not placed in rows- if they are placed in a row, the 
distance is measured between the bases of the sagittal tentacles of two succeeding polyps out 
of the same longitudinal row (intertentacular distance). Like Schultze I have retained the 
terminology of Brook, regarding the sagittal and transversal body-axis, therefore the sagittal 
axis at right angles with the colony-axis, while I call sagittal tentacles, the pair of tentacles at 
the end of the sagittal axis, and lateral tentacles, the two pairs at the end of the trans¬ 
versal axis. 
When colonies, only differing in minor points, come from different stations, I have often 
described the colonies from each station separately. In this way the great variability is made 
more evident and it is also easier to interpolate formerly described species into the series of 
variability or to unite formerly described species by way of the Siboga-material. Often I have 
availed myself of this possibility of uniting, as in my opinion again and again variants are 
described as species, since the very great width of the variability of the Antipatharia was not 
known. Once this variability given, stated by an extensive material every time coming from 
one and the same station, and knowing the fact that from many species there is only one single 
colony extant from various localities, we can easily imagine that through the influence of the 
external conditions, which doubtless will be considerable for these sessile colonies, seemingly 
