they are copiously fertile, and when sown alone reproduce their varietal form. Not only have certain forms been imparted to other Ferns, but even variegation, notably so in the Shield Fern and the Hart's-tongue. In the latter spores from a normal but variegated form were sown thickly with a plumose (or crispum form) and a branching form, and their offspring have become variegated. By sowing a muriate and a plumose Hart's-tongue together, muriate plumose varieties have also resulted.

For illustrating multiple parentage the Hart's-tongue has been selected, as the simple, strop-shaped fronds are best able to show the various departures from the normal form.

In repeating the experiment of mixed spores the varieties in each case have been limited to three or four, so that the resultant changes could be more narrowly investigated. Distinct mixtures were sown in 1887, 1888, 1889, and 1890, and the results in all the experiments established the fact that the antheridia of more than one variety have assisted in the impregnation. The varieties had conspicuously distinct characters, and in the example of 1888 the spores were gathered from a dwarf spiral form, a muricate or warty form, an undulate and a ramose one; more exactly speaking, the varieties were spirale, undulatum, muricatum, and keratoides. The parents were exhibited as well as three of their children, the latter having the names of quadriparens, Darwiniana, and echinatum. These unmistakably show on each plant the characters of the whole four parents. In the hundreds of these seedlings, as might be expected, the majority show only the characters of two parents, in a less though considerable number the characters of three, whilst a small number exhibit those of the four parents. The plants in the 1889 experiments are from a muricate, a branched, and a cup-bearing form, known as peraferens, the object being to obtain cups on a branching muricate Fern, as this was a desideratum. There was no previous example of more than one cup on a frond. In the seedlings a divided frond can be observed with cups on each division, a tasselled form with a rosette in place of an actual cup, and in another example a marginal row of small cups; and all are muricate. It is worth remarking that the seedlings from mixed spores never seem to produce any plants that exactly resemble any one variety; they are all combinations; in other words, antherozoids from a number of different antheridia are required for fertilisation. In sowing varieties of the Lady Fern I have raised the combination of five and six. This is alluded to in my paper On Prothalli.' These plants that give evidence of multiple parentage were obtained in the identical manner formulated before they had any existence. Spores require to be sown thickly to enable the prothalli to intermingle, otherwise they are only fertilised from the same prothallus. If we take the reasoning of Sir John Herschel on the doctrine of probability, and apply it to these experiments, the chances against the reasoning adopted being incorrect are as great as that of the haphazard distribution of the stars. These experiments regarding the changes in animal and vegetable life were commenced forty years ago. Bearing to some extent on this subject, experimenting on the Mimulus, a yellow variety was crossed with a spotted one, and the seedlings were spotted; later on, and further up the same stem, two blooms were this time crossed with a yellow one, but the seedlings were still spotted. The effect of the first cross had become a part of the life-history of the plant; in a second experiment the same plant was simultaneously crossed with pollen from two other varieties, and several of the seedlings are combinations of the three. It requires dexterity in crossing the Mimulus, as the pistil is as sensitive as the sensitive Mimosa. Natural changes are slow, but culturally we can accelerate that process that continues age after age. The germ once changed, the new element is retained, which becomes combined with others until the normal appearance is lost. The illustration of the Hart's-tongue shows this alteration, helped on as it were by artificial means that have accelerated the process, and these changes will continue whilst the world lasts. Affectionate respect causes tablets to be erected in memory of the departed, but age obliterates this record. It is, however, far different with the philosopher who has discovered great truths; he has erected a monument to himself more lasting than brass.' Time wears away the hardest 1891.

ΣΥ

rock, but it will require the crumbling of this world to obliterate the truths that have been taught by Charles Darwin.

3. The Ciliated Organs of the Leeches. By Professor GILSON.

It is well known that the segmental organ of the Chetopods terminates in the cœlom in the form of a funnel-shaped and ciliated structure.

In the leeches, on the contrary, though it is generally taught that there exists a similar feature, our knowledge of it is imperfect.

Several authors confess that they have not been able to detect in these worms any relation between the segmental organ and certain ciliated bodies that have been regarded by others as terminal funnels or nephridiostome.

I have undertaken with one of my pupils, Dr. Bolsius, several researches in order to resolve if possible that interesting question. I shall content myself with an extremely short account of the results we arrived at up to the present as regards the genus Nephelis.

The ciliated organs of the Nephelis are not funnel- but cup-shaped bodies, with a non-perforated bottom. The sides of the cup are composed of large bilobed cells. The bottom, on the contrary, consists only of smaller and non-ciliated cells.

This cup lies enclosed in the cavity of a large vesicle, from the sides of which it vibrates freely, being only suspended by a small number of large connective cells. The vesicle is only a dilatation of one of these non-contractile blood-vessels that represent, according to the view of Dr. Bourne, the greatly modified coelom of the leeches. It lies at a certain distance from the segmental organ, and is ordinarily separated from the same by muscles or connective cells.

The result of these observations is that the ciliated organs of the Nephelis deserve by no means the name of funnels, and that there is no anatomical connection between them and the segmental gland.

We can assert also that this gland does not open into the cœlom, at least not in certain genera of leeches, and especially the Nephelis, as it does in the wellknown case of the Chetopods.

The absence of connection between the ciliated bodies and the segmental gland seems to be a result of the profound modification the cœlom undergoes in these remarkable forms of annelids.

The terminal or coelomic part of the segmental organ is separated from the rest of the gland, and as this separation is not followed by immediate degeneration of the nephridiostome, it seems evident that the latter that is to say, the cupshaped organ-acquires at the same time a new significance and another physiological function.

As regards this new function we may propose two hypotheses which do not exclude each other:

1. The cilia cause the blood to run through the non-contractile capillaries; at least they help its motion through the coelomic system.

2. The organ is a place of proliferation of the blood-cells.-In fact the cup-shaped organ is ordinarily crowded with blood-corpuscles, the nuclei of which are often remarkable for their chromatophile power. We detected also amongst them several phases of karyocinesis.

These results, I think, are noteworthy for anyone who is interested with the position of the formation and evolution of the segmental organs and with the kidney-theory. They will soon be published in full in the Louvain Review 'La Cellule.'

4. Some Points in the Early Development of Mus musculus and Mus decumanus: the Relation of the Yolk Sac to the Decidua and the Placenta. By ARTHUR ROBINSON, M.D.

1. At the seventh day the ovum consists of a large yolk sac and a small mass of primitive epiblast which rests upon one pole of the ovum. The ovum is con

tained within a crypt in the distal wall of the uterine cavity, and the uterine epithelium is disappearing from the walls of the crypt.

2. A few hours later the primitive epiblast divides into formative epiblast and trophoblast.

3. During the latter part of the seventh day the trophoblast rapidly increases, becomes closely attached to the decidua, and pushes the formative epiblast towards the yolk sac, which becomes invaginated. The non-invaginated portions of the yolk sac lie in direct contact with the decidua, in which numerous slit-shaped blood spaces have appeared.

4. In the early part of the eighth day the walls of the ovular crypt, which sprang from the distal side of the uterine cavity, fuse with the proximal wall of that canal, and thus the crypt is converted into a closed space, and the continuity of the uterine canal is interrupted. The greater part of this space is occupied by the ovum, but at the mesometrial and anti-mesometrial ends portions of the cavity remain, and are transformed into maternal blood sinuses. The blood in the mesometrial sinus bathes the proximal end of the trophoblast, and that in the antimesometrial sinus bathes the distal end of the yolk sac. Further, by the disappearance of the inner wall of the slit-shaped spaces at the sides of the yolk sac the maternal blood is brought into direct relation with a large part of the circumference of the yolk sac, and spaces which have in the meantime appeared in the trophoblast also become filled with the same fluid.

5. During the ninth day the cœlom is formed, and the allantois, which is a solid mass of mesoblast containing no diverticulum from the alimentary canal, grows into the cœlom, but it does not become attached to the trophoblast until the eleventh day.

6. Between the ninth and the seventeenth days the decidua reflexa gradually separates from the distal wall of the uterus, and the continuity of the uterine canal is re-established. The decidua reflexa is reduced to a thin membrane, and the circulation within it ceases. When these changes are accomplished the distal part of the vitelline cavity is obliterated by the apposition of its walls, but the proximal portion remains; and by means of diverticula, which project from it into the placenta, the intimate relation of the yolk sac to the maternal blood is maintained after the circulation in the decidua reflexa has terminated.

7. The close relation of the yolk sac to the maternal blood suggests the idea that the sac itself is an important agent in the early nutrition of the embryo, and the peculiar relationship of the hypoblast to the placenta indicates the possibility that the hypoblast cells play some special part in embryonic nutrition.

5. Observations upon the Development of the Spinal Cord in Mus musculus and Mus decumanus: the Formation of the Septa and the Fissures. By ARTHUR ROBINSON, M.D.

1. At the eleventh day the spinal cord is a hollow rod of nucleated protoplasm. 2. Within a few hours the neuroblasts are differentiated.

3. On the twelfth day the formation of the grey matter commences, and the rudiments of the white columns appear.

4. The antero-lateral white columns consist of nerve fibrils derived from the neuroblasts of the cord embedded in a spongioblastic reticulum.

5. The posterior white columns are formed by the processes of the neuroblasts of the spinal ganglia.

6. The spongioblasts of the dorsal and ventral walls of the central canal are drawn out into two septa, an anterior and a posterior, which extend respectively to the ventral and dorsal surfaces of the cord.

7. The extension of the anterior septum is due to the formation of the anterior commissures and the shrinking of the central canal.

8. The extension of the posterior septum is due principally to the formation of the posterior columns, but also to the formation of the posterior commissure and the shrinking of the central canal.

9. The anterior septum does not form a complete partition between the two sides of the cord. It is traversed by the transverse fibres of the commissures.

10. The posterior septum is traversed by the transverse fibres of the posterior commissure, but it forms a complete partition between the posterior white columns.

11. There is no posterior fissure, and the posterior septum is not a septum of pia-mater, but of spongioblastic fibrils; it is, therefore, essentially a portion of the cord substance, not of its sheath.

12. The anterior fissure is formed in the usual manner, and contains a fold of pia-mater.

6. On the Innervation of the Epipodial Processes of some Nudibranchiate Mollusca. By Professor W. A. HERDMAN, D.Sc., and J. A. CLUBb.

In 1889 one of us (W. A. Herdman) read a paper at the Newcastle-on-Tyne meeting of the British Association on the structure and functions of the cerata in Nudibranchs, in which these dorso-lateral processes were regarded as being probably epipodial outgrowths. In other papers published since we have compared the conditions of these structures in various genera of Nudibranchs, and have tried to show that they are all modifications of simple lateral epipodial ridges.

The question has, however, been raised lately by Pelseneer and others as to whether the so-called epipodia of mollusca are all homologous structures, and one of the subjects of controversy now is the origin of the nerve supply in various forms, it being supposed that where the processes are innervated from the pleural ganglia they are pallial in their nature, and where supplied from the pedal ganglia they are to be regarded as outgrowths from the foot.

Consequently, it seemed to us of importance to determine afresh the origin of the nerves supplying the cerata in several different types of Nudibranchiata, especially as the results of former investigations, depending entirely, we believe, upon minute dissection, are puzzling, and to some extent contradictory. We have traced the nerves from the ganglia, by means of serial sections, in representatives of the genera Polycera, Ancula, Tritonia, Dendronotus, and Eolis, with the following results:

In Polycera quadrilineata the cerebral and pleural ganglia are completely fused to form a cerebro- pleural mass. The 'epipodial' nerves are found arising from the ventral and posterior part of this mass (i.e. distinctly from the pleural ganglia), and they run along the sides of the back to supply the ceratal ridges.

In Ancula cristata the pleural ganglia are fairly distinct from the cerebral. In a specimen cut into about 500 sections, we find in the 100th section or so from the anterior end six distinct ganglia (the cerebral, pleural, and pedal pairs) surrounding the oesophagus. A few sections further back the cerebrals disappear, and then the epipodial nerves are found arising from the dorsal edge of the pleural ganglia. The nerves soon turn posteriorly, and then give off their first branches dorsally. These branches enter the mesoderm of the body wall, and can then be traced back through over a hundred sections to the first pair of cerata, which they enter. The main nerve passes back to the remaining cerata.

In Tritonia and Dendronotus also the epipodial nerves arise from the pleural ganglia; but in Eolis (or Facelina) coronata we find that the main nerves to the cerata arise distinctly from the pedal ganglia. We have also traced in the same series of sections the ordinary pedal nerves to the foot proper, so there can be no question as to the nature of the ganglia from which the nerves arise. The epipodial nerves spring from about the middle of the pedal ganglion, rather on the dorsal surface, and, after a short course, pass through the muscular layer of the body wall and are distributed to the clumps of cerata.

But in addition to these main epipodial nerves in Eolis, we find also a nerve arising from the compound ganglionic mass, immediately ventral to the eye (probably therefore from the pleural element), which goes to the front cerata. This pleural nerve has its origin distinctly anterior to the origin of the main epipodial nerves from the pedal ganglia.

We arrive, then, at the curious result that the innervation of the ceratal processes is not the same in all these Nudibranchs. In Polycera, Ancula, Tritonia, and Dendronotus the epipodial nerves arise from pleural ganglia, or from the ventral and posterior parts of cerebro-pleural masses; while in Eolis the chief epipodial nerves are from the pedal ganglia, but there are also smaller nerves from the pleurals. In the ordinary Rhipidoglossate gastropod, such as Trochus, the epipodial ridges and processes are supplied, according to Pelseneer, by nerves arising from the dorsal part of the elongated pedal ganglia. So, judging from the nerve supply alone, it might be said that the cerata of Eolis are pedal in their nature, and homologous with the epipodial processes of Trochus, while those of Ancula and the rest are totally distinct structures of pallial origin. But these dorso-lateral processes in the various Nudibranchs are so much alike in their relations, and are connected by such series of gradations, that it is difficult to believe that they are not all homologous, and the presence of the accessory epipodial nerve in Eolis arising from the pleural ganglion suggests the possibility of another explanation, viz., that these outgrowths, starting at first as pedal structures innervated by nerves from the pedal ganglia, may have acquired, possibly as the result of having moved further up the sides of the body, a supplementary nerve supply from the adjacent integumentary nerves arising from the pleural ganglia, and this supplementary supply, while remaining subordinary in Eolis, may in the other types have gradually come to supplant the original epipodial nerves, which are now no longer found in such forms as Polycera and Ancula. This is at present only a suggestion, which, however, we hope to be able either to disprove or support by the examination of the nerves of a number of additional Nudibranchs.

7. Exhibition of a New Apparatus for Opening and Closing a Tow-Net by Electricity. By W. E. HOYLE and L. F. MASSEY.

8. Exhibition of, and Remarks upon, some Young Specimens of Echidna aculeata. By Professor W. N. PARKER, Ph.D.

The specimens are from the collection of the late Professor W. K. Parker, who received them from Dr. E. P. Ramsay, Curator of the Australian Museum, Sydney. They are much curved towards the neutral side, the snout pointing backwards, and the tail, in the older of the two stages, forwards. The younger stage measures along the dorsal curve, from the end of the snout to the tip of the tail, 12 cm., the greatest diameter of the body being 3 cm.; the corresponding measurements of the older stage are respectively 215 cm. and 6 cm. In the latter the body is covered with short scattered bristles. In both stages the snout is very similar in form to that of Ornithorhynchus, and is covered by a thick horny layer, but in other respects the specialisation characteristic of Echidna is already apparent. The gape is narrow, and extends only a short distance down the snout, and the manus, even in the younger stage, is already much larger and stronger than the pes. The tail is short and conical. There is no caruncle, or ' egg-breaker,' in the snout, such as is seen in Ornithorhynchus.

A few points in the structure of the fore part of the head in the older stage were described. The mouth has the narrow and tubular form seen in the adult, and the long tongue has a horny tip. The glands in relation with the mouth and nose are very numerous. There is no trace of any teeth-rudiments, and in many other respects the structure of the head shows extreme specialisation. Jacobson's organ is large and highly developed; a well marked 'turbinal' is present in it.

1 J.e. dorsal to the foot, whether there is a distinct pallium present or not.