22 ARVID R. MOLANDER, ALCYONACEA. 
allow us to presume this way of evolution. The membranous Cornularids show the di- 
rection of the evolution, but no direct transitions between Cornulariidae and Alcyoniidae 
are found. 
The fam. Xeniidae differs from this series of evolution. 'The colonies have a fleshy 
build, the spicules are few in number and constitute a type which is distinguished 
from the type of spicules of Cornulariidae and Alcyoniidae. The fleshy consistency of the 
colony compensates for the slight development of spicules for supporting the colony. 
Within Xemiidae we also find dimorphismus. 
Within Alcyomiidae the evolution has followed different directions. It appears to 
me that Nidalia is the most primitive genus. The spicules are of uniform shape, mostly 
spindles and rods, and the same spicules occur both in anthocodiae and in the coenenchym 
of the colonies. The colonies are shaped like cones or rollers and are mostly undivided. 
Sinularia differs from this genus in the peculiar, club-shaped spicules in the bark of the 
upper part of the colony; the colony is divided into shaft and end plate, the tentacles 
have more than one line of pinnulae. Only the dorsal mesenterial filament generally 
occurs. Dimorphismus may occur. According to KUKENTHAL (1906), the superficial 
entodermal canal system is possibly connected with the non-existence of siphonozooids. 
These canals will be rudimental coelenterons of the siphonozooids. Further the develop- 
ment within the family follows two lines. 1) The colonies are without siphonozooids. 
They are well developed and generally divided into lobes. The spicules are spindles 
and rods, which are to a great extent mixed with double stars (dumb-bells), and 
crosses. In the bark of coenenchym the latter are most numerous (Alcyonium). 
2) The colonies have dimorphous polyps. The colonies are either cylindrical, unim- 
portant, or not divided (Acrophytum, Lobophytum), or they have the shape of mushroom 
(Anthomastus, Sarcophytum, Nidaliopsis); the end-plate is often divided (Antho- 
mastus). The spicules in the polyps are spindles and rods, which partly pass over into 
rollers with girdles, double stars, etc. In the bark of the upper part of the colony the 
spindles and rods are mixed with a number of double stars etc., especially within the 
most developed genera Anthomastus and Lobophytum. Within Lobophytum and Antho- 
mastus the dimorphismus 1s most developed with numerous siphonozooids, and within 
Anthomastus the few autozooids are large and the siphonozooids only contain generative 
tissue. The dimorphismus has been developed within both the fam. Xeniidae and Alcyo- 
midae. But the dimorphismus of both the families has been developed independently 
of each other, and the dimorphic line of Alcyoniidae has not been developed from 
Xenidae. 
The position of Sympodium is doubtful. The peculiar growth of colonies can not 
be considered as a characteristic of the nature of Cornulariidae, and the genus shows the 
nature of the Alcyoniidae (p. 19, 35). The armature is reduced, and the spicules are 
small and slightly thorned. The shape of the spicules partly reminds one of that of 
Alcyonium. Short rods, mixed with crosses, stars and irregular calcareous bodies are 
found. The canal system mostly consists of cell-strings, which in the case of Sympo- 
dium catenatum also have a vertical distribution and are placed not only at the base 
of the polyps. At the base we also find larger horizontal canals. The partly lamellated, 
