406 Scientific Reviews. 



has been the cause of introducing among the polypi many natural 

 productions^ bearing a certain resemblance to them in external ap- 

 pearance, but which differ from them entirely in almost every other 

 respect, nay in some cases do not even belong to the animal king- 

 dom. In proof of this, it is sufficient to refer to the diversity of 

 natural objects which have been brought together in the chaotic 

 genus Alcyonium. The author, therefore, proposes to take the ani- 

 mal itself as the basis of his arrangement, while he uses the charac- 

 ters derived from the nature of the stem only in establishing the 

 subordinate divisions. 



According to this method, all polypi are arranged into two great 

 divisions, the Exoaria and Endoaria, which differ chiefly in their 

 mode of propagation. In the first, or Exoaria, the parts destined 

 for this purpose grow from the external surface : they consist either 

 of solitary naked ova or germs, or of small capsules, each of which 

 bursts and discharges a number of ova or young animals already 

 formed. In the Endoaria the ova are produced in the interior of 

 the body, and are either conveyed outwards by means of oviducts 

 which open by separate orifices, or they are discharged by the 

 mouth. The singular mode of propagation by sprouts is not takeii 

 into consideration here, because it belongs in common to all the po- 

 lypi, and consequently can afford no ground of distinction between 

 them. 



The division Exoaria comprehends three families : 1. the Hydrae ; 

 2. Corynea, consisting of the genera Sertularia, Tubularia, and 

 Coryna, also characterized by this mode of propagation ; 3. Mille>- 

 pora. The families belonging to the Endoaria are, 4. Alcyonea, 

 the Polypes Tubiferes of Lamarck, of which the genera are the 

 Anthelia and Xenia of Savigny, Ammothea of Lamarck and Lobu- 

 laria, distinguished from each other chiefly by the mode of connec- 

 tion of their tubes, and by the circumstance of the animals being 

 capable or incapable of withdrawing themselves into the tube ; 5. 

 Tubipora, consisting of but one genus ; 6. Corallia, whose genera 

 Corallium, Gorgonia, Isis, Antipathes, are characterized by pecu*. 

 liarities in the structure of the stem ; 7* Pennatulae, the genera of 

 which are almost the same as those given by Cuvier ; 8. Zoantha, 

 whose two genera Cornularia and Zoanthus, are distinguished from 

 each other by the number of their tentacula ; 9. Madrepora : the 

 author considers the animals of this family to be nearly allied to 

 tJie Actiniae, from which, however, they differ in propagating by 

 ova or germs, while the Actiniae produce the young alive. The 

 M. denudata (Cavolinia rosea, Schweig.) seems to form the transi- 

 tion between them. The genera of Madrepora enumerated are 

 those of Lamarck. 



The Actiniae are treated of at greater length, and in some de- 

 gree form the subject of a distinct part of the work. All the 

 known species are described with many interesting observations of 

 the author relative to their history, which are partly new, partly 



