disengage themselves during the roasting of the auriferous matters, carrying the metal with them, M. Varinski made a series of experiments, in which the results confirmed entirely this supposition.

1. Different quantities (1, 2, and 3 poods) of argentiferous martial pyrites not containing any gold, and in which the quantity of silver was determined by assays carefully made, was mixed with 24 zol. of powder of gold, prepared by a chemical method. In each experiment this mixture was roasted upon a tile in a muffle, afterwards it was submitted to scorification and cupellation, and they obtained as a result 14, 16, 20, 22, 26, 37 a 41 per cent. less than the quantity they mixed with the pyrites.

2. Lastly, to demonstrate that the sulphur was not the only substance which in volatilizing itself carried the gold with it, but that there are others which give rise to the same phenomenon, he made a second series of experiments; therefore 24 zolotnicks of powder of gold purified chemically, were mixed with 1, 2, and 3 poods of an ore of iron, containing the oxide of this metal, and water and carbonic acid, but in which there was not either gold, or silver, or sulphur. This mixture was similarly roasted upon a tile in a muffle, and was treated by scorification and cupellation; they obtained by this means from 7 to 23 per cent. of gold less than they had put into the mixture. 3. 24 zol. of gold were mixed with different quantities of pyrites, and of the same sort of ore of iron spoken of before; this mixture was roasted upon a tile in a muffle and treated by scorification and cupellation. The loss of gold was from 4, 6, 8 to 24 per cent.

4. 24 zol. of gold were mixed with a pood of volcanic sulphur. This mixture was submitted to a treatment identically the same as the preceding, but the loss of gold was not more than 4 per cent, probably because the powder of gold smelted promptly, and the vapors of sulphur could not have had time to elevate themselves in particles, as in those which were carried on more slowly.

5. Lastly, to ascertain if this loss of gold operated in roasting in open vases equally as in closed ones, they made the following assays: 24 zol. of gold with the pyrites and the ore of iron were treated by scorification and cupellation, with the preliminary roasting in open and in closed vases. The result of this roasting demonstrated, that when it took place in closed vases the loss of gold was from 16 to 25 per cent., and in open vases from 23 to 29 per cent.

They concluded from these proofs that gold and silver in the metallic state, and which are mixed with substances containing bases which are susceptible of volatilization by the action of heat, volatilizes with them although in different proportions. That the loss affects particularly the metal which is found is the least quantity in the roasted mixture. That the loss is the more considerable as the roasting is the more complete. That in the assays by the dry way for gold and silver, it ought as it appears, to renounce the preliminary roasting admitted until now as a rule, and scorify immediately without roasting, because all of the assays above executed on a small as on a large scale, prove invariably that the roasting produces a loss of the precious metals. When treating a considerable quantity of ore, and of products containing silver, and above all, gold, if it is possible, avoid the roasting. Lastly, it would be desirable that this object should be examined with great care, by means of assays executed in different localities and under different circumstances, upon different substances containing gold and silver, for the purpose of ascertaining if the loss

gave rise to the supposition that at the moment of evaporation of the sulphur and of the other volatile substances, the portions of the precious metals are carried off, mechanically mixed with thin vapors, the same as lead volatilizes itself when a current of gas passes over its surface. Gold is raised with the greatest case when it is disseminated in such minute particles, and it would be retained in this form in the air, with the same ease as when it is suspended in the liquids.

The chief of the mining staff, who was then inspecting the works of Siberia, was cognizant of the experiments of Varinski, offering evident proof that the loss of precious metals resulted from the roasting of substances which contained them, and the idea occurred to him that there would be a similar loss during the roasting of the matts of the furnace at the works of Altai. On his arrival at the works of Bamarul, he made experiments upon the roasting of the matts in the furnaces and in the open air. The results of these experiments are shown in the following table, proving that the more complete the roasting is, the greater is the loss of metal, particularly of gold. In seven different roastings the loss per cent. of the parts of the metals, was:

of silver 2 11 3 11 23 25 28 of gold 43 43 32 42 62 100 100 The soot recovered from the chimney of the furnace for roasting, submitted in great quantity to repeated assays, for the extraction of gold and silver, gave of a zol. of silver, without a trace of gold for each pood of soot. It is admitted that they obtain annually from the works of K. N. in Altai 120,000 poods of the matt of the furnace, containing nearly 120 poods of auriferous silver, which contains about 3 poods of gold, and they concluded that there were lost in the roasting about 2 poods of gold and about 11 poods of silver; a very considerable loss for these mines. It was determined then to avoid, if possible, the roasting of all the substances which contain gold and silver.

The chief of the mining staff being returned to Ecatharineburg, made there anew the experiments upon the schlichs of auriferous minerals of Berezofsk, which were submitted to amalgamation raw and roasted; it was demonstrated by these assays:

1. That amalgamation obtained the whole of the gold contained in the schlichs, because the quantity of gold furnished by it was precisely the same as that determined by the assays by acids.

2. That the roasting was objectionable for the amalgamating substances, producing a loss of metal.

3. That this loss, particularly remarkable for gold in matts, affected it preferably to silver; the loss of the latter was elevated to 10 a 32 per cent., and that of the first into 80 per cent.; it was said that after all, probably we might attribute this loss during the roasting of the schlichs in the reverbatory furnace, to the disengagement of vapors and of gas, which carried off the attenuated particles of the precious metals.

To be convinced more completely that the volatile substances

disengage themselves during the roasting of the auriferous matters, carrying the metal with them, M. Varinski made a series of experiments, in which the results confirmed entirely this supposition.

1. Different quantities (1, 2, and 3 poods) of argentiferous martial pyrites not containing any gold, and in which the quantity of silver was determined by assays carefully made, was mixed with 24 zol. of powder of gold, prepared by a chemical method. In each experiment this mixture was roasted upon a tile in a muffle, afterwards it was submitted to scorification and cupellation, and they obtained as a result 14, 16, 20, 22, 26, 37 a 41 per cent. less than the quantity they mixed with the pyrites.

2. Lastly, to demonstrate that the sulphur was not the only substance which in volatilizing itself carried the gold with it, but that there are others which give rise to the same phenomenon, he made a second series of experiments; therefore 24 zolotnicks of powder of gold purified chemically, were mixed with 1, 2, and 3 poods of an ore of iron, containing the oxide of this metal, and water and carbonic acid, but in which there was not either gold, or silver, or sulphur. This mixture was similarly roasted upon a tile in a muffle, and was treated by scorification and cupellation; they obtained by this means from 7 to 23 per cent. of gold less than they had put into the mixture. 3. 24 zol. of gold were mixed with different quantities of pyrites, and of the same sort of ore of iron spoken of before; this mixture was roasted upon a tile in a muffle and treated by scorification and cupellation. The loss of gold was from 4, 6, 8 to 24 per cent.

4. 24 zol. of gold were mixed with a pood of volcanic sulphur. This mixture was submitted to a treatment identically the same as the preceding, but the loss of gold was not more than 4 per cent, probably because the powder of gold smelted promptly, and the vapors of sulphur could not have had time to elevate themselves in particles, as in those which were carried on more slowly.

5. Lastly, to ascertain if this loss of gold operated in roasting in open vases equally as in closed ones, they made the following assays: 24 zol. of gold with the pyrites and the ore of iron were treated by scorification and cupellation, with the preliminary roasting in open and in closed vases. The result of this roasting demonstrated, that when it took place in closed vases the loss of gold was from 16 to 25 per cent., and in open vases from 23 to 29 per cent.

They concluded from these proofs that gold and silver in the metallic state, and which are mixed with substances containing bases which are susceptible of volatilization by the action of heat, volatilizes with them although in different proportions. That the loss affects particularly the metal which is found is the least quantity in the roasted mixture. That the loss is the more considerable as the roasting is the more complete. That in the assays by the dry way for gold and silver, it ought as it appears, to renounce the preliminary roasting admitted until now as a rule, and scorify immediately without roasting, because all of the assays above executed on a small as on a large scale, prove invariably that the roasting produces a loss of the precious metals. When treating a considerable quantity of ore, and of products containing silver, and above all, gold, if it is possible, avoid the roasting. Lastly, it would be desirable that this object should be examined with great care, by means of assays executed in different localities and under different circumstances, upon different substances containing gold and silver, for the purpose of ascertaining if the loss

specimen of a Trilobite (Paradoxides Tessini or Harlani) which vas obtained by him from the slate quarry at Braintree on the 9th ugust last. He also exhibited a cast of a specimen of the same ecies, which from the character of the rock was undoubtedly ained from the same ledge, and which was purchased by Mr. rancis Alger at the breaking up of the old Columbian Museum of Boston, some twenty-five years ago, and was originally presented to that Museum by some one residing in this vicinity. Mr. Alger's specimen is in a sharp angular prismatic mass of rock, having all the appearances of having been broken from the rocks in place, and certainly was not a boulder.

From the existence of this specimen, and also from the frequent discovery of fragments of trilobites in the erratic rocks on George's island, Geologists were prepared for the discovery of them in some of the ledges of this neighborhood, but no one ever thought of looking among the pinched-up and metamorphic slates between the Quincy and Braintree sienite hills for any fossils, until they were actually disclosed by the quarrying operations of the Messrs. Haywood at Braintree, and one of our members, Peter Wainwright, Esq., recognized them as trilobites and as subjects of great scientific interest, and called the attention of professed Geologists to the locality.

About five years ago, Mr. Eliphas Haywood first observed these fossils on opening his stone quarry for the purpose of obtaining underpinning and ballast stones. Without knowing their nature, he still looked upon them as interesting curiosities, and laid aside the specimens which have lately been brought before the Society.

He showed them to Mr. Wainwright, who at once recognized them as trilobites and brought them to Boston for the inspection of Geologists, and presented two specimens to our associate, Prof. Wm. B. Rogers, to whom the Society is indebted for the first notice of these remarkable fossils, so important in the determination of our geognostic horizon. A few days after Prof. Roger's visit to the quarry, Dr. Jackson, by invitation of Mr. Wainwright, visited it and made a minute examination of all the geological phenomena which it presents, and obtained specimens of the trilobites through the kindness of Mr. Haywood, and by search at the quarry in company with Mr. Wainwright. Two specimens were obtained, one entire, which is 8 inches long and 4 inches wide.

The other, of which only the head and half the body was obtained, is 6 inches wide, and its hood is 7 inches across by the base of the head; hence the length of this specimen must have been 12 inches at least, which is about the size of the largest specimens of the Paradoxides Tessini discovered in Sweden. The smaller individual has 21 articulations, but none in the tail beyond the lateral appendages, and in this respect differs

ART. III.-THE SILURIAN SYSTEM; ITS BASE IN MASSACHUSETTS.*

The Secretary read the following communication from Prof. Wm. B. Rogers:

So far as I have yet explored the quarry in the Quincy and Braintree belt, containing the remarkable fossil Trilobites to which I referred at the preceding meeting of the Society, I find that they belong chiefly, if not altogether, to one species, which on the authority of Agassiz, as well as my own comparison with Barrande's descriptions and figures, is undoubtedly a Paradoxide. Of its specific affinities I will not now speak further than to remark that the specimens agree more closely with Barrande's Par. spinosus than with any other form.

As the genus Paradoxides is peculiar to the lowest of the paleozoic rocks in Bohemia, Sweden and Great Britain, marking the Primordial division of Barrande and the Lingular flags of the British survey, we will probably be called upon to place the fossiliferous belt of Quincy and Braintree at or near the horizon of our lowest paleozoic group, that is to say, somewhere about the level of the Primal rocks, the Potsdam sandstone and the Protozoic sandstone of Owen, containing Dikelocephalus in Wisconsin and Minnesota.

Thus for the first time we are furnished with data for fixing conclusively the paleozoic age of any portion of this tract of ancient and highly altered sediments, and what is more, for defining in regard to this region the very base of the paleozoic column, and that too by the same fossil inscriptions which mark it in various parts of the Old World.

One of the most curious facts relating to the trilobite of the Quincy and Braintree belt, is its seeming identity with the Paradoxides Harlani described by Green in his monograph of North American trilobites. This description, which is quite imperfect, was made out from a specimen of unknown locality, procured some twenty-five years ago, through Dr. Harlan, from the collection of our well-known mineralogist, Mr. Francis Alger. The identity is, I think, established by the comparison of a nearly complete specimen of the Braintree fossil with the cast of P. Harlani, taken from Mr. Alger's specimen, the original never having been returned. Considering the perfect agreement in lithological character of the matrix as described by Green, with that of the Quincy fossils, and the immediate recognition of this agreement in mineral features by Mr. Alger, on seeing my Quincy specimens, we can hardly doubt that the original specimen of P. Harlani came either directly, or through the drift scattered in the vicinity, from the same fossiliferous belt.

Dr. Jackson presented to the Society a cast of a very perfect.

*Boston Society of Natural History.