multiplied, and especially the dependence of this number, 330, on the greater or less fusibility of the slags, and on the quality of the iron, has to be further examined.

As a résumé, we may in the mean time adopt, as the caloric absorbed by the reduction of the ores, etc., and the fusion of the iron for gray forge iron, No. 3,

§ 12. Caloric absorbed by the fusion of the slag, the decomposition of the limestone, etc.-The slags have very different degrees of fusibility. Many years ago, Sefström and Berthier ascertained that bisilicates and even trisilicates of lime, magnesia, and alumina, are more fusible than the protosilicates, and generally the most fusible silicates correspond to compositions near to bisilicates.* It is on this account that the bisilicate formula is aimed at in all cases where the presence of sulphur or some analogous motive does not demand an excess of lime in the charges. And from this circumstance we see that not only the caloric absorbed by the slag, but that of the iron yielded, must vary, as above indicated, with the chemical constitution of the slags. The difference in the fusibility of slags has been proved by Plattner also, by comparing it with various alloys of gold and platinum. But if fusibility varies with the composition of the slags, so must the total caloric. The earthy singulosilicates, which are little fusible, must take more caloric than the bisilicates of these bases, or than the silicates containing a certain proportion of alkalies, and of

* See Note II. Appendix.

the oxides of iron and manganese. Hence, a diversity of results has been obtained by experiments on slags as by those on cast-iron.

For a very irony slag from a cupola, MM. Mineray and Résal found 336 calories.

For a slag, approximating to sesquisilicate of lime and of magnesia, M. Rinman found 441 calories, and for another 430. For a glassy slag from charcoal furnace, having manganese in its composition, and approximating to a bisilicate, M. Gillot found 370 to 380 calories.

MM. Dulait and Boulanger found, for a slag of No. 3 iron, 433 calories, and for one from gray iron for foundry, 492.

These two latter slags, like the most of slags coming from blast furnaces working with coke, are generally nearly singulosilicates. The first certainly contained oxide of iron.

Lastly, M. Vathaire found, for a slag from a coke furnace yielding No. 3, 350 calories; and Mr. Bell found even 572 calories, but considers this number as a too high determination. From what precedes, we see that slags which are bisilicates and contain manganese, do not retain more than 370 to 400 calories in flowing from the furnaces, whilst sesquisilicates retain about 450; and that singulosilicates may take as much as 500 calories when they contain neither iron nor manganese, but, on the other hand, a high proportion of earthy bases.

Again, Mr. Bell admits 550 cals. for the total caloric of slags of Cleveland No. 3 iron, by reason of their strong proportion of lime and alumina.

In any case, the slags always retain more caloric than the iron. They have both a higher specific heat and a higher latent heat. This latter is as high as 120 cals., according to

Rinman, for sesquisilicates, whilst that of cast-iron is only 46. But we see from the diversity of results, that in order to have an exact estimate of the caloric absorbed, we should have to make special experiments on each slag.

We have more exact experiments on the caloric absorbed in the decomposition of limestone. MM. Favre and Silbermann found the number 373.5 calories for calcspar, and 360-6 for arragonite. Thus, in this case again, the molecular state has a certain influence on the caloricity, and it cannot be affirmed that every limestone, crystalline or amorphous, dense or porous, requires the same sum of caloric for its decomposition. Still, we may admit 373.5 for our present purposes.

We have now to estimate the caloric absorbed by the vaporization and decomposition of water. For the evaporation we shall adopt Regnault's number, 606.5 calories.

For the decomposition of water, we have 29,003 calories per lb. of hydrogen set free. This is the caloric produced by the combustion of hydrogen to steam. The result, therefore, is 29,003 3222 calories lb. of water.

9

=

per

Let us remark in conclusion, that though the fusion of the slags absorbs caloric, the combination of silica and the bases probably disengages a certain amount of caloric which we cannot estimate.

§ 13. Sensible heat carried off by the Gases.-The caloric

* Annales de physique et de chimie, 3d series, t. xxxvii.

1 lb. hydrogen produces 34,462 calories, the steam being condensed to 00. If water remained in the state of vapor at 0°, we must deduct 9+ 606.5 =5458.5. Hence the figure 29,003 above given. Compare Bell, section xli. Tr.

carried off by the gases is easily calculated if we know their composition and their temperature.

It is only necessary to consider separately each constituent of the gaseous mixture. Taking the specific heats determined by Regnault we have per lb. and for each degree Centigrade:

and we find further on, that, according to the mean composition of the gases of blast furnaces (coke), the specific heat is as a general average very nearly 0.237.

§ 14. Caloric lost by radiation from the walls of furnace, etc. -The caloric thus lost is composed of several parts. There is the heat carried off by the cooling water, which is easily determined; there is the heat dispersed by radiation from the walls; that which the air carries off in its currents past the walls, and that which passes into the base of the furnace by conduction. These two latter cannot be determined, but we may attempt to determine what is lost by radiation.

Mr. Bell made experiments on this subject on a blast furnace on the Wear. He used an oblong vessel of copper holding about nine quarts of water, every side of which except that put to the furnace was cased with flannel and with wood, with interposed thin strata of air.

By applying this uncovered side to different parts of the wall of the furnace, Mr. Bell ascertained the caloric given off per unit of surface, and hence for the whole surface of the furnace. In this way he found—

For the Wear furnace per lb. of iron

And for the caloric carried off by the water of the twyres, 10,150 lb. of water heated to 90.16 Centigrade

And to this must be added an allowance for caloric carried off by the air currents, and that lost in the foundations. And thus the number 300 or even 400 calories may be adopted for this source of loss of caloric.

§ 15. Determination of the Caloric received by a Blast Furnace.-Let us now return to the caloric produced in the interior of the furnace. Neglecting the caloric resulting from the combination of the elements constituting pig-iron and the slags, the caloric produced is derived solely from the transformation of carbon into a certain mixture of CO2 and

CO.

This caloric may be calculated, either by deduction from the analysis of the gases, or by considering separately the zone of combustion near the twyre and the zone of reduction in which CO is transferred into CO2.

Let us first apply a knowledge of the analysis of the gases. Making use of the notation of § 7, we have y = the weight

of CO, and my the weight of CO2, and hence

3

y=

7

the carbon in CO, and my the carbon in CO2. But the

3 11

carbonic acid contains b of carbon derived from the limestone,

therefore the carbonic acid produced by combustion only con

· ·

tains (my —b) of carbon.