Nov. 29, 1S83] 



NA TURE 



115 



whilst in the female Physophorid;c the origin of the single ovum 

 is different (in the endocodon). As soon as the homogeny of the 

 two layers of the Ccelenterata with the two primitive layers of 

 the higher Metazoa became evident, the question arose whether 

 the germinal cells of the Metazoa generally were of ectodermal 

 or endodermal origin, and a large number of observers attempted 

 to settle the question offhand by investigating the process of 

 development of the germinal cells in some one Ccclenterate. 

 Each assumed that his particular results must hold good for the 

 entire group, and as the results were conflicting — the place of firs! 

 appearance of the germinal cells lying as is well known in some 

 Ccelenterates in the ectoderm and in others in the endoderm — much 

 confusion arose. At this period, E. van Beneden's memoir 

 appeared which, on the strength of the conditions occurring in a 

 Hydractinia, a Campanularia, and a Clava, started the theory 

 that the germ layers were themselves sexually differentiated, the 

 female elements arising from the endoderm and the male from 

 the ectoderm, and that in the union of a derivative of each layer 

 lay the essence of impregnation, the neces-ai7 precm-sor of repro - 

 duction. This brilliant conception was soon shown by further 

 observation to be erroneous, and as Prof. Weismann points out 

 it was from the first not in accordance with the phenomena of 

 parthenogenesis. As the next important phase iu the question 

 came the attempt of the brothers Hertwig to prove that the 

 Coilenterata belong to two distinct stocks, the one consisting of 

 the An'hozoa and Scyphomed sa;, in which the germinal cells 

 are derived from the endoderm (Endocarpa;), and the other of 

 the Hydromedusa: and Ctenophora, in which they originate from 

 the ectoderm (Ectocarps). If this position bs correct, and, as 

 will be seen in the sequel, one of the mo 4 startling of the con- 

 clusions arrived at in the present work is that, notivithstanding 

 all the apparent evidence to the contrary, it probably is so in 

 re.al'ty, then the important principle of inheritance and continuity 

 in development in the germ layers receives a strong support, of 

 which with regard to the gonad elements it seemed in great need. 

 Prof. Weismann was led to undertake the present prolonged 

 researches by his observing that in certain of the Hydromedus^e 

 the germinal cells originate, not in the sexual individuals them- 

 selves nor even in the blastojtyles that support them, but in 

 the C(Kao=arc of the colony, in the common parenchym of 

 tlie stem and its branches, and that this occurs not only in 

 the case of the female but also in some instances in that of 

 the male germinal cells. T.he existence of ovicells of ccenosarcal 

 origin had been previously observed by Quatrefages, F. E. 

 Schultze, Fraipont, and others, but these elements had not been 

 reco.;ni-ed as the sole source of supply of the female gonophores 

 with ova. E. van Beiieden further hai observed the origin of 

 the egg-cells in Hydractinia, in that part of the blastostyle which 

 sub equenlly becomes evaginated to form the gonophore. 

 Kleinenherg published his account of his discovery of the migra- 

 tion of the egg-cells of Eudendrium from the ectoderm into the 

 endoderm and in the opposite direction ju>t before Weismann 

 had arrived at a similar conclusion and had found in his pre- 

 parations eggcelU in the act of boring through the basement 

 me.iibrane with one half lying in the ectoderm and the other in 

 the endo,l;rm. The establishmeit of the fact that migration of 

 the .sexual cells of a m>st remarkable character in the many forms 

 in which he has proved it to occur is a constant phenomenon, 

 the history of its details, and the discussion of the phylogenetic 

 origin and general biological bearings of the curious phenomena 

 presented by it, form the most important features of the present 

 work . 



The author as more c mvenient adopts — instead of AUman's 

 terms, pha'icrocodmic gonoplioreand adeloodonic gonophore — 

 "medusa" and "medusoid gonophore" respeclively. He applies 

 the latter term to all gonophores, n it becoming free medusa;, in the 

 walls of which any traces, however rudimentary, can be detected 

 of the three layer.s, viz. the inner and outer ectoderm layers and the 

 intervening endjderm lamella — of which the wall of the bell of 

 the medusa is composed. He uses the term sporophore for 

 those gonoph:)re sacs in which no indication of anything beyond 

 a single layer of ectoderm and endoderm can be disc ivered. 



A structure which assumes great importance in the hi-tory of 

 the wanderings of the ovicells is the duplicature of ectoderm, 

 which grows inwards at the summit of the simple sac-like bud 

 out of which a medusa i? formed, depressing the endoderm 

 lanella and forming the hollow of the bell. It is necessary thit 

 this embryonic organ or mass of cells, observed by so many in- 

 vestigators, should receive a special name, and it is termed 

 "endocodon." 



It is pointed out that each hydranth of a colony does not con- 

 sist alone of that part containing the stomach and bearing the 

 tentacles and hypostome, but also of a stem-shaped portion, 

 which is developed at the same time with it out of the same 

 bud. This region is termed the '" hydrocope," and is included 

 in the hydranth, the remaining region of which is the "hydro- 



iiiiti 



Fig. I — Diagram of a bud of a medusa or medusoid gonopliore — Glk, endo- 

 codon ; Glfi, sub-umbrella space : Entl, primitive endoderm lamella : 

 ttkz, primitive germ cells ; cut, endoderm ; e/it, ectcderm. 



cephalis." The hydrocope corresponds to the region in Tubu- 

 laria which AUman terms hydrocaulus, but not to the whole 

 system of stems and branches in an arborescent colony. In such 

 colonies the production of buds is entirely confined to the hydro- 

 cope and its counterpart in the blastostyle, the " gonocope." 

 In the Tubularince it is necessary to distinguish amongst the 



Fig. 2. — Diagram of a primary. Hhy, and lateral, Shy, hydranth of Euden- 

 drium : He, fiydrocephalis : H, neck ; Canib, cambium zjne; Km, zone 

 of gemmation: i/(?/, hydroc-npe : Sa, lateral branch; Bist, blastostyle; 

 Ccp, gonocope; ^ph, sporophore. 



hydranths of a stock the "principal" from the "lateral" 

 hydranths. The principal hydranths are those v\hich remain 

 permanently at the extremities of the stems or branches 

 throughout the growth of the stock by lateral buddinj;. In the 

 arborescent stocks of the Tubularida: the first hydranth sprung 

 from the egg remains penuanently at the extremity of the 



