24 



PROF. AGASSIZ S 



These animals which were extremely numerous 

 in former geological ages, agree in the mode of 

 growth of their plates, with the young of that star- 

 fish called Comatula, as it has been observed by 

 Thompson. This diagram (Plate I, fig A.) seems 

 to represent a large animal but it is only highly 

 magnified, the natural size of it being only half an 

 inch long. Nevertheless we distinguish in it the 

 articulated stem with joints. We have the crown 

 above with its solid plates. We have the dividing 

 arm arising from it. We have the surrounding 

 tentacles contributing to seijze its prey and bring it 

 to the mouth, and the movable tentacles or suck- 

 ers along the inner side of the branched rays, 

 which this animal moves as the others use their 

 suckers. In addition to these, there are gradually 

 more tentacles coming out, and the body grows 

 larger, till it is freed from its stem in its perfect 

 condition. The star-fishes which do not rest upon 

 a stem and which uo not branch, resemble less 

 those fossils than the types of them in which the 

 rays are more numerous and in which the rays 

 branch (Plate I, fig. B.) But even in common star- 

 fishes in their earliest condition (Plates IV, and X, 

 fig. A-), we have an arrangement of the solid parts 

 which resemble more closely the arrangement of 

 the solid parts of Crinoids (Plate X, fig. B.) than 

 the arrangements of parts in the full grown star- 

 fish (Plate VIII, fig. A.) Compare the solid plate 

 in the young starfish with the solid plates of the 

 fullgrown animal. In the young we have a star in 

 in which five large plates seem to alternate with five 

 others, ten of them forming the principle mass of 

 the body. 



[PLATE X COMPARISON OF THE CALCAREOUS 



NET WORKS OF STAR FISHES, WITH 



THE SOLID PLATES OF GRINOIDS.] 



Pentacrinus, here the five plates which surrop.nd 

 the mouth, and those alternating with them, will 

 form the five rays, and so on with successive little 

 plates in all the genera. 



["PLATE VII STAR FISHER CRI>'OIDP.J 



If we take the Pentacrinus (Plate XV, fig. A) we 

 observe above the stem a crown, in which five large 

 plates, forming the cup, alternate with five smaller 

 ones. In Apiocrinus, the larger plates constitute a 

 hollow cup and above them alternating with them, 

 there are others (Plate XV, fig. B) upon which the 

 branching arms rest. In Encrinus crown and arms 

 are not so widely separated and seem to form still 

 an undivided cavity, as in the genera of Plate 

 XVII. (Plate VII, fig. D). Everywhere the same 

 arrangement exists, so that on a diagram the same 

 drawing would answer for the crinoids and the 

 common star-fishes indiscriminately. Here 

 (Plate X, fig. A) is the central network of the 

 common star-fish, corresponding to the |stera of 



In PUite X , tig. B, we have the corresponding 

 parts in a diagram of a crinoid, answering precise- 

 ly in position, number, and mode of growth, the 

 so^id frame of the starfishes. In all we find a plan 

 which is uniform, whether we observe such ani- 

 mals in which the young are provided with a 

 stem, or those in which the stem does not appear 

 at all. Even if we go back to the young echini, 

 which seem to differ so much from the starfishes, 

 we have an identical number of primitive suck- 

 ers, namely, five, (Plate XI, fig. F) which do not 

 give rise to a pentagonal body until the flattened 

 disc assumes a more spherical form (Fig. G). So 

 that there is a most intimate agreement between 

 the different growths of the embryo. 



All these data upon the embryonic changes of 

 Echinoderms are very fragmentary, as I have al- 

 ready remarked ; nevertheless, with these incom- 

 plete series of observations, it can be shown, as I 

 think I have done, that these embryonic forms 

 agree intimately with those which occupy a higher 

 rank in the class, and that they resemble also the 

 form of those which existed in former geological 

 ages. 



Would these data afford the means of now in- 



