Cottus is at the bottom of the stream among the insects and minute water-plants. Johnson and Willoughby represent the female as much distended with spawn, which they say "she collects in little lumps on her breast, where it is covered with a black membrane until it is hatched."

It is a great pity that naturalists should make assertions which they cannot substantiate. I have had a great many of the fish under close examination, in an aquarium set apart for the study of their habits, and cannot recall a single circumstance which would justify the assertion that the female takes care of the eggs.

Not long since Mr. Sarel of Brighton, presented me with four specimens of the Cottus gobio or miller's thumb. These I placed in an aquarium holding thirty-six gallons of water, continually running. On the morning of the 4th of March I was surprised to find that a large light coloured specimen had become thin, and a small black fish had taken up a position on the top of a stone, near the side of the tank. On closer inspection I found that the light fish had laid her eggs sideways against the glass, at the top of the stone, which was about two inches high, and in so convenient a position, that I rightly anticipated I should be able to watch the whole process and progress of the hatching.

The little black fish rested upon the eggs, which were about forty in number, and kept fanning them with his large pectoral fins incessantly, while with ferocious eyes he watched the actions of the other occupants of the aquarium. After fanning the cluster by a downward movement, as if he had calculated upon their being deposited upon the ground (as they ought to have been), he seemed suddenly to realize their unnatural position, and reversed the action of his fins, thus drawing the current of water upwards through the fixed mass.

This mistake frequently occurred daily, and was as often rectified. Amongst the eggs were several channels, through which the water passed freely, in fact every pellet was exposed to a current in the same manner in which the eggs of the whelk are arranged. While the male fish thus kept watch, he never partook of a particle of food, and any attempt to supply him was resented; worm after worm was offered to him, but he merely took it in his mouth and removed it wriggling to the most distant corner of the tank. On placing an iron bar in the water, near him, to remove a pebble, he seized it savagely, and held it so firmly that I lifted him out of the water nearly a foot, and kept him suspended for about a minute.

A live shrimp was introduced near the spot where the eggs lay;

he seized it by the middle, and took it away like a dog would a rat, afterwards returning to his position to recommence fanning the eggs more vigorously and even savagely, as if to make up for lost time. If any foreign substance dropped from the fernery above him, he would seize it and rapidly carry it off to the other side of the aquarium.

Four days after the dark Cottus had commenced his labour of love, the eggs began to assume a yellowish tint. The fish seemed to be in great distress at the cool impertinence of the sticklebacks, who came curiously watching his proceedings. Not only had he to guard against these strangers, but the parent who laid the eggs was in a lean condition, and seemed to be watching her opportunity to devour the spawn.

Sixteen days after the eggs were laid or deposited, they appeared to be speckled. On using a lens I discovered to my delight that every egg contained a live fish, and these little spots were eyes. The sentinel now began to get furiously excited, and fanned away day and night without ceasing; first downwards, and then upwards. He rushed round the tank and attacked every other occupant without the slightest provocation, and again rapidly returned to his duties. On the seventeenth day a beautiful rose-coloured blush momentarily suffused each egg, and disappeared, and every now and then there appeared a rapid half turn of an object within the egg, then I had the satisfaction of discovering that all the eggs were hatching quickly.

Thirty days after their being deposited, the crimson blush appeared and disappeared seventy-five times a minute, and thirtyfive days after a small fish issued from most of the eggs. I was called away from home on the day the last of the fish were hatched, and on my return the tank was dry, some accident having caused a leakage. Enough, however, was noticed to settled the question as to whether the male or female hatches, or broods over the eggs. This conclusion might be very well attained by analogy. We find that all fish we are able to study are possessed of two kinds of socalled roe; the hard and the soft-the former belonging to the female, and the latter to the male. When the female sheds her roe or spawn, the male fish broods over it and deposits the soft roe or milt upon it, totally regardless of the presence or absence of the female, who generally departs as soon as she has performed her mission. Whether in ordinary cases any considerable time is taken up in the fulfilment of the duties of the male, I know not, but that this devoted male fish not only hovered over the eggs for more than

a month, but after the little fish had escaped he still continued at work, fanning the egg-shells, and was seen thus occupied on the day the tank ran dry.

I may mention one curious incident to show the desire of a Cottus to exercise parental control. On the occasion of one of the small fish emerging from its egg, the parent chased it to the bottom of the tank, and taking it gently in his mouth, replaced it in the cluster. This, however, was not repeated, for the young ones emerged very rapidly from their tiny prisons, from the first hour of hatching, as if the struggles of each assisted the others to escape.

ARCHEOLOGIA.

A VERY interesting ROMAN SEPULCHRAL MONUMENT has recently been found at COLCHESTER, the Camulodunum of the Romans, on the property of Mr. George Joslin, jun., which is on the site of the extensive cemetery of the ancient city, which lay on the road towards Londinium (London). It is a sculptured stone, the upper part of which has a figure, about three feet high, of a Roman soldier, of unusually good workmanship; beneath which the following inscription presents itself, in four lines, and in large and well-cut letters :

M FAVON MF POL FACI

LIS LEG XX VERECVND

VS ET NOVICVS LIB POSV

ERVNTH'S'E

Which may be read, Marcus Favonius Marci filius Pollio Facilis (centurio), legionis xx. Verecundus et Novicus liberti posuerunt. Hic situs est. The figure, as just stated, is well sculptured, and is especially interesting for the details of the costume. The centurion holds in his left hand a dagger, and he carries a second dagger at his right side. In his right hand he holds the rod, or vine branch, the centurion's badge. His legs are encased in greaves, and he wears a richly-ornamented girdle. We may remark that the twentieth legion had its permanent station at Deva (Chester). Camulodunum, or, as it was more commonly called, Colonia, appears to have been, like Uriconium (Wroxeter), a favourite place for retired veterans.

On the eastern side of Colchester, on the road leading to Mersey, Roman sepulchral remains have also been discovered within the last month, but we have not yet any exact description of them.

The CONGRESS of the BRITISH ARCHEOLOGICAL ASSOCIATION at St.

Albans will take place during the week beginning on Monday, the second of August, with Lord Lytton for president. The programme will, we believe, soon be published, and we have heard of several interesting papers already preparing for reading during the week.

Some early BARROWS between Sidmouth and Honiton, in Devonshire, in one of which was found a remarkable interesting DRINKING CUP, made of KIMMERIDGE COAL, and evidently turned on the lathe, It is about three inches and a half in height, and nearly three inches in diameter, and of a rather unusual form.

A quantity of ROMAN COINS have been dug up on SALISBURY PLAIN by some labourers employed at work there. They lay at a depth of about two feet, and had been deposited in a coarse earthen jar. In number they were between eighteen and nineteen hundred, of small brass, belonging nearly all to the Constantine family, and ranging from Licinius (A.D. 307-324) through the period when the imperial throne was occupied by that family. Mr. Roach Smith, who has communicated some remarks upon these coins to the Numismatic Society, points out the great interest in the exergual lettering of these coins. It is well known that at this period of the Roman empire in the west, considerable mints were established in some of the greater towns, and the place at which the coin was struck was indicated by certain letters in the exergue. Thus P.LON (percussa Londinio) indicated that the coin was from the Mint at London. Others marked PLC and PLN have been ascribed to the same mint, but there existed some doubt whether these belonged to the London Mint, or to that of Lugdunum (Lyons), where also there was an important establishment for issuing money. Mr. Roach Smith, in the remarks just alluded to, has given very satisfactory reasons, derived from a comparison of these coins found on Salisbury Plain, for considering that these coins marked PLC and PLN could not have come from the same mint which produced the coins marked PLON, and that therefore they are not from the Mint of Londinium. By far the greater part of these coins found on Salisbury Plain belong to the Mints of Treveris (Treves), Londinium (London), Lugdunum (Lyons), and Arelatum (Arles).

A very interesting discovery has lately been made at LUTON, NEAR CHATHAM, for the particulars respecting which we are indebted to Mr. Humphrey Wickham, of Strood, by whose prompt decision the remains, accidentally brought to light, were preserved. Luton, as the name indicates, lies below the chalk hill, in a very sheltered site. During excavations for brick-earth and buildings, the work

men came upon walls, formed chiefly of flints, which, at some remote period, had been broken down to within about three feet of the ground on which they stood. Further digging revealed the fact that it had been a walled family burial-place. It contained the following objects, which had apparently been inclosed in a wooden chest a bronze jug, of elegant shape: a bronze bowl, bearing the maker's name, AFRICAN (Africanus); the bronze hasp of a chest; an iron lamp-stand; an iron implement upwards of three feet in length, which has, perhaps, been a weeding spud; a square glass bottle and a portion of a ribbed glass basin; and four dishes of what is usually called Samian ware.

T. W.

CORRESPONDENCE.

LUNAR ECLIPSES.-ETA ARGÛS.-THE EARTH SHADOW.

We have received the following in a letter from Mr. Thomas D. Smeaton, of Robe, South Australia :

In "Chambers' "Handbook of Astronomy," it is noted that during a total Eclipse of the Moon, March 19, 1848, the Moon was so readily seen that many persons doubted the reality of the eclipse. On September 24, 1866, the same phenomenon was remarked in South Australia. On the other hand, the eclipses of June 15, 1620, and May 18, 1761, were remarkable for the complete disappearance of the Moon. Is it more than a coincidence that in the first two instances the eclipse took place near the Equinoxes, and in the last two near the Solstices? And if this were the rule, could any fair explanation be founded upon it?

In Guillemin's "The Heavens," the English Annotators seem to have overlooked the statement (p. 334) that Eta Argûs now presents an appearance of unusual magnificence, and surpasses Canopus in brilliancy. So far from this being true, it is to the naked eye almost lost in the nebula that accompanies it. (Proctor, 1866, continues to mark it of 2nd mag. So also Dietrichsen and Hannay, 1869.)

In the clear sky of South Australia the Shadow of the Earth is often seen at sunset mounting in the eastern sky as a low dark sector. The phenomenon of rays proceeding from a point opposite to the Sun is also often observed, and on one occasion I saw a rainbow partially encircling these luminous rays, like the tire and spokes of a wheel. The whole upon a dark thunder-cloud as back ground.