88 



JOHNS HOPKINS 



[No. 88. 



fessor Haeckel mistook for male gonophores some long, spindle-shaped 

 gynophores filled with yolk, but into which the egg nucleus had not yet 

 passed, or from which it had been forced out by pressure (several such 

 gonophores were seen). It seems proper here to mention the fact that an 

 examination of the sexual organs of forty or fifty specimens of Physalia has 

 shown that they were all males. The structures described by Haeckel as 

 the female gonophores of Physalia are not gonophores, but swimbells, as 

 is shown by an examination of living expanded specimens. 



Concerning the female gonophores we are able to speak with certainty. 

 Since the present investigation has shown that the so-called " Polyovone 

 Gynophores " show no trace of medusoid structure, and are therefore strictly 

 speaking not gonophores at all, but merely pouches containing ova, the 

 term " polyovone gynophore" has been replaced throughout this paper by 

 the term "egg-pouch." The so-called "Monovone Gynophores" are true 

 gonophores, and are known throughout this paper merely as female gono- 

 phores. The remarkable varieties of " Monovone " and " Polyovone Gyno- 

 phores," mentioned by Haeckel, were not observed at all, and it seems 

 probable that at least many of these varieties were merely different stages 

 in the development of the female gonophores. 



The structure of the mature gouophore is so complicated that it will best 

 be understood if the development of the gonophore be traced from its simple 

 beginnings to its final complex form. The earliest stage shown is in Fig. 6, 

 in which a small bud, A, is represented as arising from the distal end of a 

 large egg pouch, B, which also bears a well developed gouophore containing 

 an egg. The egg pouch itself communicates with the digestive cavity of 

 the stem or gonostyle, S. The ectoderm has disappeared over most of the 

 external surface of the bud. The latter is merely a protuberance blind at 

 its distal end, opening into the egg-pouch at its proximal end and lined 

 with endoderm, which is continuous with that which lines the egg-pouch 

 and the digestive cavity. 



The next stage is shown in Fig. 1. As before, the gonophore arises from 

 an egg-pouch ; however, in this case there are no clearly marked ova in the 

 latter, though it and the stem of the gonophore are filled with yolk, Y. The 

 egg-pouch is connected with the stem of the gonodendron near the point 

 where the pedicel, P, of the gonophore arises, but at a slightly lower level 

 than that represented in the figure. The invagination at the end of the 

 bud is caused by the pushing in of a solid mass of ectoderm, B N; this is 

 the hell nucleus ("glockenkern"), and the cavity of the invagination, filled 

 with the mass of ectoderm, represents the umbrella cavity. 



Fig. 2 shows a stage still more advanced. The egg pouch is connected to 

 the stem at a different level from that shown in the figure. The egg cells 

 in the pouch are quite numerous and in all stages of development. It is 

 clearly seen from the sections that the egg cells are endoderm cells which 

 have grown in size, protruding from the simple layer of endoderm which 

 originally lined the egg pouch into the cavity of the latter. In many cases 

 the egg cells in growing have torn the endoderm cells from the supporting 

 lamella and have carried them up along their sides, so that the endoderm 

 cells come to lie between the large egg spheres. Usually the egg cells are 

 found in their original position in the endoderm layer, i. e. on the support- 

 ing lamella; some, however, have separated from the lamella and are free 

 in the cavity of the egg pouch or in that of the pedicel, P, of the gono- 

 phore. In the specimen figured two egg cells lie in the cavity of the 

 pedicel below the invaginated endoderm, while one has crowded into the 

 narrow space between the two layers of cathammal endoderm, K. En., on 

 the side of the bell cavity. The ectoderm in the bell cavity, which in the 

 previous stage was a solid mass, now lines the cavity down to the bottom 

 of the invagination, where it turns up on each side and unites with the 

 ectodermal mass which fills the mouth of the invagination. The endoderm 

 and supporting lamella also bend outward at the bottom of the invagina- 

 tion. This feature varies greatly in different individuals; in some it is 

 scarcely seen at all, in others it is quite marked. There can be little doubt 

 that the structure represents the manubrium or proboscis of the ancestral 

 medusa. 



The next stage in the development of the gonophore, and the one most 

 interesting, namely, that of the migration of the egg from the egg pouch 

 into its final position, was the stage most difficult to obtain. On careful 

 search, however, several such stages were found. Fig. 3 shows an egg from 

 the egg pouch just breaking through the endoderm and supporting lamella. 

 The mass of yolk still tills the pedicel while the distal part of the egg, Y, 

 has pushed into its final resting place, carrying the ectoderm of the manu- 



brium before it and flattening it out against the ectoderm of the subum- 

 brella. There is a wide break, R, in the endoderm and supporting lamella 

 through which the nucleus of the egg is just entering. Gradually, however, 

 as the egg gets into its permanent position this break closes and we reach 

 the condition shown in Fig. 5. The egg is now in its permanent position 

 and the endoderm and lamella are closing in behind it, T, to shut it off 

 from the pedicel and egg pouch. A second smaller egg is shown just at the 

 break, which, if one may judge from other similar cases, will be broken 

 down and absorbed by the larger egg now in position. 



Fig. 4, a stage corresponding to Fig. 3, is nevertheless different from it, in 

 that the gonophore does not arise out of the egg pouch but from the stem 

 of the gonodendron. However, from the same opening in the stem a small 

 egg pouch, B, arises. In this case it is impossible to say whether the egg 

 originated in the small egg pouch and then migrated into its present posi- 

 tion in the gonophore, or whether it originated above or below its present 

 level and then migrated through the stem to its present position. We have 

 in no case been able to find an egg originating in the stem or in'any other 

 place than the egg pouch, and we consider it extremely improbable that 

 they do arise in any other position. Nor have we been able to find any 

 egg migrating through the stem, so that in the case of Fig. 4 it is probable 

 that the egg did originate in the adjacent egg pouch and then migrated into 

 its present position. The condition of the egg pouch in this case indicates 

 that it is undergoing degeneration, and considering the fact that the egg 

 pouch is present in all cases where the gonophore is young, and that as this 

 becomes more and more developed the egg pouch becomes smaller or is 

 absent altogether, we hold it probable that the eggs not only originate in 

 the egg pouch and afterward migrate into the gonophore, where alone they 

 mature, but that the egg pouch also furnishes nutriment to the egg in the 

 gonophore, and that as the latter grows larger the former is absorbed by it. 



In Fig. 6 the gonophore is still further developed. The edges of the 

 ruptured endoderm and lamella which, in the previous figure, had closely 

 approached and had almost shut off the egg from the egg pouch and the 

 digestive cavity, have now grown together, completely closing the break by 

 which the egg entered. The egg does not completely fill the gonophore, a 

 small cavity, U, being left at its distal end, which is the remnant of the 

 umbrella cavity. At this point the two layers of the ectoderm immediately 

 surrounding the egg are separated ; the inner layer (nearest the egg), M. EC., 

 is the ectoderm of the manubrium, the outer layer, S. EC., is the ectoderm 

 of the subumbrella; outside this are two layers of endoderm, the catham- 

 mal lamellae, closely pressed together, and in which no remnant of radi- 

 ating canals may be found, but which do show traces of a ring cana,l at the 

 distal end of the gonophore where the endoderm folds on itself. Covering 

 the outside of the gonophore and filling the mouth of the bell is a layer of 

 ectoderm which is continuous with that covering the stem. In fact, all the 

 ectoderm of the gonophore is in one continuous layer, as appears in the 

 figures. This layer was at first invaginated (primary bell invagination), 

 and then subsequently evaginated by the pushing in of the. egg. In the 

 same way the endoderm is in one continuous layer. The endoderm of the 

 stem, gouostyle and egg pouch being continued through the pedicel of the 

 gonophore, and along its side to its distal end, and there folded back on 

 itself by the primary invagination. The endoderm just below the egg and 

 separating it from the pedicel shows a slight upward protuberance, M. En., 

 on each side. This represents the nianubrial endoderm, and is the begin- 

 ning of a fold of endoderm which pushes up on all sides of the egg and 

 nearly surrounds it, leaving only a small place where the egg still remains 

 in contact with the ectoderm. Fig. 7, M. En., shows these endoderm folds 

 more largely developed than in Fig. 6 ; there is a slight cavity (continuous 

 with the digestive cavity) running up between the folds. 



Fig. S is a gonophore almost mature, though many reach a size two or 

 three times as great as this. The endoderm folds have grown up far around 

 the egg, and the cavity which they enclosed in the former stage is no longer 

 seen. At their distal ends the folds are imperfectly cleft into a number of 

 finger-like processes which arise from the proximal part of the folds as the 

 fingers do from the hand, only in this case, to carry out the simile, the 

 fingers are connected by a web; the egg is surrounded by these as a ball 

 which is held in the hand is surrounded by the palm and fingers. In the 

 figure the folds are constricted in several places, which is due to the fact 

 that several of these finger-like processes are cut across in the section. 

 These endoderm folds" and finger-like processes have a marked glandular 

 appearance; and there can be no doubt that the cells of these endoderm 



