beauty of that romantic spot. Two of the commonest Hong-Kong ferns belong to the real shield-ferns, A. molle, Sw., and A. unitum, Sw. Both are alike, and apparently resemble the English male fern, A. filix-mas, but if the student examine them closely he will find a connection in the veins at the sinus, or bend of each little division, or lobe. Of the veinlets that branch regularly up each segment, a pair meet and terminate at every bend or cut. This meeting of veinlets constitutes the main difference between two of the largest subdivisions of this enormous genus, viz. Lastrea and Nephrodium. A. molle, Sw., and A. unitum, Sw., are both about two feet high, and grow everywhere in the hills and in the town. A. molle is a light green, soft and downy ; A. unitum is darker and more shiny, the former has very few very small sori, often found only close to the rachis. In A. unitum the fructification is dark-coloured, denselycrowded, and closer to the margin of the lobes.

Gen. XI. MENISCIUM, Schreb.

M. simplex, Hk., is not uncommon at the Peak, and is easily distinguished by the very marked raised veins, laid as it were like net-work all over the under side of the frond, which is about 4 or 5 inches long, entire, and very finely pointed, having two half-lobes at the base more or less detached from the main-stem. The long delicate apex is often half the length of the entire leaf. The fertile portion differs materially in being much longer, much narrower, and more upright than the sterile, and closely packed with the fructification, which is brown and destitute of covering.

Gen. XII. POLYPODIUM, Linn.

Every polypody does not resemble our old friend in England, whose yellow buttons of sori and favourite haunts on old trees and ruined walls render it familiar to us all. The technical distinction is the absence of indusium (often overlooked in aspidium, when the indusium is sometimes obsolete or lost).

There is no polypody common to this neighbourhood. The searcher in the hills might perhaps be rewarded by finding P. adnascens, Sw.-a little fern covered with furry down on the under side, and round sori, or P. lingua, Sw., rather larger, with less down or tomentum and sori at further intervals. In both ferns the frond is entire.

(To be continued.)

AN Excavation carried on by the German Government is said, in the Times, to have been in process near Schladebach, with the object of obtaining further information as to the increase of underground temperature. At 1392 metres, the depth reached at the beginning of this year, and believed to be the lowest yet attained by boring, the temperature was 49° C.

THE VARIATION AND ABNORMAL DEVELOPMENT OF THE MOLLUSCA.

L

PART II.

IMNEA GLUTINOSA. This remarkable species belongs, not to Limnæa proper, but the genus Amphipeplea of Nilsson, and is easily known from Limnæa by the fact, that when alive the mantle covers the shell, and in this way and in the texture of the shell it is related to Physa fontinalis. This species is rare and local, and does not vary so much as some of the allied species. In the district we are now dealing with, it has been recorded as living on Barnes Common; my brother has specimens from Deal, Minster, and Sittingbourne, and I have found it abundant but local in the St. Nicholas Marshes, where it is easily seen on the leaves of Nuphar luteum and other plants. These St. Nicholas Marsh specimens are rather light in colour, and it was there I found a most curious and interesting monstrosity or variety, which has much the relation to the type of glutinosa that L. involuta has to L. peregra.

This specimen has the spire very short, and sunken, but slightly raised at the apex; the body whorl is swollen above, and the top of the shell appears nearly flat. Should this form turn up in any other localities it might be called monst. intortum, but as long as we have only a single specimen I think it is better unnamed.

Limnæa peregra. This is the commonest and most variable of our freshwater mollusca.

I have taken a very globose form, probably var. ovata, in the Regent's Canal, where the water is stagnant and there is very little weed. (I record the kind of situation when possible, as one often finds that facts, seemingly of no account, are afterwards valuable in drawing conclusions as to the origin and use of variations.)

A specimen from Chislehurst is slightly decollated, specimens from Bromley and Eltham are somewhat thinner than usual, but my thinnest specimen was taken by my brother at St. Nicholas Marsh. I have an exceeding thick and apparently semi-fossil shell from Barnes, and from what has been said above concerning the Barnes shell it seems improbable that it could have recently lived on the spot. This species occurs fossil at Crayford, but the specimens do not differ from those now living. I have a succineashaped specimen taken in a well at Farnborough, it is possibly the variety succinaformis of Jeffreys. A monstrosity from Kew Gardens has a wide and deep umbilicus. A shell I found in a ditch close to Walmer Castle has a rather long spire with a fairly deep suture, and has a number of confluent whitish bands all of which are below the periphery. (It is remarkable that whenever bands are abnormally developed in the genera Limnæa and Physa, they are

usually, as far as my experience goes, below the periphery.)

There is a pond at Bromley where L. peregra occurs. The specimens are very variable in shape, and are covered with some kind of growth which måkes them appear almost black outside, and greyish within. They are infested with a parasite, which my brother tells me is a very beautiful microscopical object, but I have not yet examined any microscopically myself. Some of these shells have a very expanded lip, and belong to the var. labiosa, Jeff., and in some cases the lip of the shell is even reflected upon itself, so that the mouth presents a rounded edge formed by the inner surface of the lip which by the reflection becomes outermost.

L. auricularia. The var. acuta is recorded for Kent. Specimens from Regent's Park have a rather long spire and are pale in colour.

L. stagnalis. Var. fragilis has been taken in Middlesex and Surrey. Monst. scalariforme, Chislehurst (S.C.C. and T.D.A.C.). Monst. decollatum,

Barnes.

A variety (which might conveniently be called expansa)* lives in a small pond at Bromley, where the type form does not occur. It has a short spire, the body whorl is large and expanded, and the mouth of the shell wide; the length of the spire is about of the total length of the shell, which is somewhat less than an inch and a half. The only weed I noticed in the pond was Lemna minor. A specimen of L. stagnalis taken at Deal is slender in shape, light in colour, and has a shallow suture. Shells from a small pond on Chislehurst Common, on the contrary, are dark in colour, and have a deep suture (two specimens being actually scalariform, as stated above), and these shells are a good deal smaller than the type form; some specimens, apparently full grown, being about two-thirds of an inch in length, although others are much larger. Should these be found elsewhere, var. elegantula would perhaps be a suitable name for them. Another pond, also on Chislehurst Common, produces quite a different variety. This form has a very shallow suture, and is not unlike some varieties of L. palustris. It is never so large as the typical stagnalis, the usual length being little more than an inch. Why these two ponds, only a few hundred yards distant from one another, should produce two forms so totally distinct, I cannot imagine; the soil appears to be the same, and the only difference I can detect is that the one having the first variety is in the open, is very small, the Limnææ being very crowded, especially in the summer, when the pond is almost dried up, and the food plants are Potamogeton and Ranunculus. The second pond is partly under the shade of a chestnut tree; it is much larger and not so

Unless it should prove identical with the variety lacustris of Moquin-Tandon, or the variety fucinensis of Paulucci, both of which it resembles, in some respects at any rate.

+ Mr. Taylor, of Leeds, is of opinion that this form does not differ from the var. botanica of Clessin.

crowded, and the principal food-plants are Anacharis and Callitriche.

From a third pond on Chislehurst Common I took a specimen having whitish bands below the periphery, like the Walmer L. peregra. Specimens from Grove Park have a pointed and slender spire, and the lip tinged with pink.

I will now venture to say a few words about the naming of varieties. There are some who would name every variation that can be defined, and others go to the other extreme, and will have nothing to do with varietal names. To me it seems that as every, or nearly every species has one or more marked variations from what is called the "type," and likewise a number of less definite varieties connecting one form with another, it is extremely difficult to find an intermediate course between giving names, or ignoring them altogether. Suppose, for instance, in the present paper, I had refused to accept any varietal names, but had written out a full description of each variety, what a deal of space would have been uselessly expended; and the difference between the space taken up by the names of described forms and descriptions of undescribed ones, is well shown above. "But," some reader will exclaim, "you will be naming almost every shell that passes through your hands, and my memory will be burdened with thousands of names which are really of no importance; and although a species is a tolerably fixed point, varieties are infinite, and naming them can only end in confusion." These are sentiments I have heard expressed.

I would reply to this, that the varietal names are principally of use to specialists, for it is manifest that a general student of conchology is scarce able to study all the variations of all the species; but to one studying any particular genus or species, the varietal names are most useful and almost indispensable. In proof of this I will ask, How many specialists have been obliged to make varieties, and how many have done without them? I fancy the former are in the majority. And some think that a species is a fixed point. To these I would recommend the study of Pisidium. Others say that varieties are of no importance; I contend that they are of the greatest importance, but I cannot enter here into the reasons why they are obvious enough to most people.

As for burdening the memory, this is an equally good argument against bringing forward any information whatever, and against names in general. Lastly, I would remind those who do not already know it, that all names, generic, specific, or varietal, are merely a matter of convenience, and although it is open to anyone to propose names, no one is obliged to accept them unless he finds it convenient to do so. Some people, however, seem to think that a sort of mysterious change comes over a specimen when it has a name tacked on to it, and its value goes up 95 per cent. I think it would not be a bad idea if at periodical

It

times the opinion of the conchological world were invited as to the advisability of accepting names proposed in the interval, and these names rejected or accepted as it might be found convenient. But the worst of it is that people will follow their own opinion in such matters, whatever others say. may be asked here, what are the convenient names, and which the inconvenient ones? In my opinion the convenient ones are those which apply to varieties which are widely distributed, and consequently have often to be referred to, or for any other reason often spoken of, and which express as well as a single word can express, the leading feature of the form in question. Constantly recurring monstrosities I would place in the same category. Names which I would reject are those which refer to a single specimen, no similar one having been found; but of course, should the form be found afterwards in several places, the

tincta, St. Nicholas Marsh (S. C. Cockerell). In these specimens the suture is exceedingly shallow. Var. albida, Minster (S. C. C.). I have taken it at Sandwich. Monst. decollatum, Barnes. Monst. carinatum. Mr. J. W. Taylor described this from a specimen found by my brother in a pond at Bromley. I have searched this pond, but have not found another carinate specimen, although the type is common enough; but while looking for carinatum I found another monstrosity, equally curious. It was a turretid specimen, which may be described as follows:-Shell about half an inch in length, whorls 5, spire turretid, suture deep, last whorl more than half the total length of the shell, and flattened at the sides, instead of being rounded, as in a typical specimen. The upper whorls are somewhat eroded.

This specimen reminds one of a monstrosity (imperiale, I think it is called), of Buccinum undatum,

name might again be taken up. Also those which do not clearly express what is meant in the description given for them, as var. major, "shell larger than type," without saying how large, or var. clongata, "shell elongated," when no length is stated, and also all those named after the discoverer, or anyone else, such as Clausilia rugosa, var. Everetti, "shell smaller than type." So in the names I have suggested above, I do not for a moment propose that they, shall be used on the strength of one or two specimens, but merely propose them as useful names, should it be found at any future period convenient to use them.

Limnæa palustris. Var. conica has been taken in the Thames. Some which I took in a ditch near the river at Putney, Mr. Kenneth McKean considers to belong to this variety, but although they are lighter than usual, they are hardly greyish-white, and the suture is, if anything shallower than usual. Var.

Fig. 123.-Limnæa stagnalis, var. elegantula." Chisle

hurst.

Fig. 124.-Limnæa palustris, monst. “"turritum."

and gives the idea of a specimen which had become telescopic, and had got partially shut up. If it is to be named, it might be called monst. turritum.” Monst. globosum, Taylor, Enfield, one specimen (S. C. Cockerell). Although this was described as a variety, I have no doubt whatever that it is really a monstrosity, and my brother is of the same opinion. My brother has taken two specimens of a very interesting variety, which has the whorls compressed,. the suture perfectly flat, and the periphery bluntly angulated; in fact, this form bears exactly the same. relation to L. palustris type, as Buccinum acuminatum bears to the typical undatum. (The correspondence between the abnormalities of L. palustris and B. undatum is very curious, especially as the two species are so distantly related.)

Kreglinger mentions a var. turrita, which may possibly be identical with the form here described.