64 
angles are 90°, in the other one nearly go°. Length of branches 0.5 — 4.5 cm.; mutual distance 
a few mm. to 1 cm. Spines 60 y, at a distance of 330 y ; 5 longitudinal rows and straight 
quincunx; form as fig. 47 a. Polyp-remains very low; tentacles in three pairs. Interpolypar 
distance 1.5 — 1.75 mm.; length of tentacles sometimes 1.3 mm., but usually shorter. — 
Station 95. Irregularly curved stem of 6 cm., regularly tapering; 7 straight or slightly curved 
branches at variable distances (± 0.5 cm.), 1.5 cm. in length, and at an angle of 6o° to 90°, 
not lying in a plane. Spines ditto as in fig. 47«, but with concave sides; length 45 y ; distance 
— : 95 g-; 5 longitudinal rows with a quincunx here and there. Spines sometimes doubled or 
irregularly shaped. 
In the tabel opposite this page I have put together all described specimens and the species, 
which I would identify with them and with Pallas’ Antipcithes dichotoma , as far as I could 
give numerical data with the help of former descriptions and figures. The letters added to the station- 
numbers in the first column indicate the specimens in their order of description. While the review 
of characters could be complete for the Siboga-specimens, this was not always possible for the 
formerly described species, as either the text, or the figures left smaller or larger gaps in the 
descriptions. As far as possible I have deduced the characters and the numerical data from 
the text as well as from the figures, both of which methods did not always give the same 
results; often only approximately computed values could be given through defective figures, the 
missing of the enlargements-data, or because of a vague text. Now already I will point 
out that most species are represented by some fragmentary colonies only, often without polyps 
even, and that only very rarely complete colonies were available as a base of the descriptions, 
and moreover only one single specimen in most cases (cf. the former habitat, following the 
diagnosis furtheron, where the number of colonies is given for each species, between brackets). 
In the case of the Siboga-material, in some cases two or even more complete colonies 
were available. To simplify the review I am going to give, the various species and stations 
are indicated by their roman ciphers. 
Number of longitudinal rows of spines. In most cases these rows are 4 in 
number. The deviations, which occur, are the following: 
4— 5 rows by XII; XIVp- XVII; XX; XXIX; XXXII. 
5 rows by II; III; VII; VIII; XV; XVI; XIX; XXI/3; XXII; XXIII; XXIV; 
XXXVI; XXXVII. 
5— 6 rows by XXI«; XXVIII; XXXVII. 
6 rows by X ; XI; XXV. 
6— 7 rows by IX; XXVII. 
These deviations are not very great, especially if one bears in mind that the number of 
rows increases on the older parts of the colony, as is demonstrated by various Siboga-specimens 
and as some authors emphatically mention in their descriptions. Much more than six rows do 
not occur. The large number of species with 5 rows might be explained by the fact that the 
authors perhaps gave the middle value of the number of rows on different parts of the axis 
(this is not the case in the Siboga-specimens with 5 rows), or by also counting the rows of 
spines the bases of which were invisible from one side of the axis while their tops were visible. 
