There are, von Baer considers, four main types--(1) the peripheral or radiate type, (2) the longitudinal type, (3) the massive or molluscan type, (4) the vertebrate type. The radiate type is shown by discoid infusoria, by medusae, by starfish and their allies. The longitudinal type characterises such genera as _Vibrio_, _Filaria_, _Gordius_, and all the annulate animals. Mollusca, rotifers, polyzoa, and such infusoria as are not included in types (1) and (2) belong to the massive type, in which the body and its parts form rounded masses. The longitudinal type is predominantly "animal," the massive type predominantly "plastic" (vegetative). The vertebrate type has both the "animal" and the "plastic" organs highly developed. In the symmetrical arrangement of the animal parts it resembles the longitudinal type; its plastic parts with their asymmetrical arrangement and rounded shape belong to the massive type. These types of von Baer inevitably recall the "Embranchements" of Cuvier, with which they more or less coincide. It seems that von Baer arrived at his types (from the study of adult structure) independently of Cuvier, though the priority of publication rests with Cuvier.[174] Now it is clear that the development of the individual, which is essentially an _Ausbildung_, a differentiation, is directly comparable with the grade-differentiation of forms within the type. And just as the type rules all its varied modifications, so does the development of the individual take place always within the bounds imposed by type. This is von Baer's chief contribution to the theory of embryonic relationships--the law that "the type of organisation determines the manner of development" (p. xxii.). Development is not merely from the general to the special--there are at least four distinct "general" types, from which the special is developed. The type is fixed in the very earliest stages of development--the embryo of a Vertebrate is from the very beginning a Vertebrate (p. 220), and it shows at no time any agreement in total organisation with any Invertebrate. The types are independent of one another; differentiation and development follow a different course in each of them. Not but what some analogies can be found between the very earliest stages of embryos of different type. Thus vertebrate and annulate embryos agree in certain points at the time of the formation of the primitive streak. And in the earliest stage of all, the egg-stage, there is probably agreement between all the types. In eggs with yolk, whether vertebrate or annulate, there is always a separation into an animal and a plastic layer. It seems, too, as if a hollow sphere were a constant stage in the development of all animals (pp. 224, 258). Apart from these analogies, development takes an entirely independent course in each of the four main types, and no embryo of one of the higher types repeats in its development the peculiar organisation of any adult of the lower types.