Feb. 26, 1885] 



NATURE 



401 



1200 feel, a gradient of 41 5 feet to the experiments in granite, 

 and of 39 feet to those in slate. 



Taking the experiments of Mr. Fox in eight mines, varying 

 in depth from 1100 to 2100 feet, the mean of the experiments 

 made in the rock gave a gradient of 43'6 feet per degree, or for 

 the mean of the two observers we have a gradient of 43 feet per 

 degree. 



For the foreign mines, in the absence of fuller data, and 



especially failing in information of the depth of the station 



beneath the surface, which in the hilly district of Freiberg and 



Hungary introduces an element of great uncertainty, it is imposs- 



rive at any safe conclusion. 



Artesian Wells and Borings. — This class of observations pre- 

 : 1 1 -i much more uniform, and whereas the mines obser- 

 vations were made, the one in crystalline, and the other in 

 unaltered Palaeozoic rocks, the wells are, with few exceptions, 

 either in the softer and less coherent rocks of Cretaceous, Juras- 

 sic, and Triassic age, which are much more permeable, and, as 

 a rule, much less disturbed. 



The causes of interference are mainly reduced to pressure on 

 the instruments and convection currents. The early experiments, 

 where no precautions were taken against these, are, with few 

 exceptions, unreliable, and must be rejected. The larger the 

 bore-hole, the greater the risk of convection-currents, and Prof. 

 Everett has shown that in many cases of deep and large artesian 

 borings, the water which lodges in them is reduced to a nearly 

 uniform temperature throughout the whole depth by the action 

 of these currents. In the deep boring at Sperenberg, before the 

 introduction of plugs to stop these currents, it was found that 

 the temperature near the top of the bore was rendered 4°'5 F. 

 too high, and at the bottom, at a depth of 3390 feet, 4°'6, if not 

 6° 7, too high by the currents. 



Taking the bore-holes in which the A-ater does not overflow, 

 and where, owing to the precautions against these sources, such as 

 those of Kentish Town, Richmond, Grenelle, Sperenberg, Pregny, 

 "I, we get a mean gradient of 51 '9 feet per degree. 



Overflowing artesian wells should, if we were sure of all the 

 conditions, give the best and most certain results. Taking those 

 where the volume of water is large, and the observations made 

 by competent ob.-ervers, as in the case of the wells of Grenelle, 

 Tours, Rochefort, Mondorff, Minden, and others, we obtain a 

 mean of 50/2 feet, or, taking the two sets of wells, of 51 feet per 

 degree. 



The author, how ever, points out a source of possible error in 

 those wells, arising from a peculiarity of tubage which requires 

 investigation, and owing to which he thinks the water may suffer 

 a loss of heat in ascending to the surface. 



With respect to the extra-European wells, more particulars 

 are required. It may be observed, however, that the wells in 

 the Sahara Desert, which were made by an experienced 

 engineer accustomed to such observations, the mean of eleven 

 overflowing wells, at depths of from 200 to 400 feet, gave 36 

 feet per degree. 



Tunnels. — For the Mont Cenis Tunnel, allowing for the con- 

 vexity of the surface, Prof. Everett estimates the gradient at 79 

 feet, and for the St. Gothard, 82 feet per degree. But Dr. Stapff 

 found in the granite at the north end of the tunnel a much 

 greater heat and more rapid gradient, for which there seemed 

 no obvious explanation. Though this axis of the Alps is of late 

 Tertiary date, the author points out that it cannot be due to the 

 protrusion of the granite, as the Swiss geologists have shown 

 that the granite was in its present relative position and solidified 

 before the elevation of this last main axis of the Alps, and he 

 suggests that the higher temperature may be a residue of the 

 heat caused by the intense lateral pressure and crushing of the 

 rocks which accompanied that elevation, for in the crushing of 

 a rigid material such as rock almost the entire mechanical work 

 reappears as heat. 



Conductivity of the Rocks. Effects of Saturation and Imbi- 

 bition. — Some of the apvarent discrepancies in the thermometric 

 gradients are no doubt due to differences in the conductivity of 

 the rocks. Applying the valuable determinations of Profs. 

 Herschel and Lebour to the groups of strata characterising 

 the several classes of observations, the following results are 

 obtained :— 



Mean Mean 



conductivity 

 h 



(1) Carboniferous strata "00488 



rystalline and schistose rocks ... '00546 

 ;i friassic and Cretaceous strata ... '00235 



275 

 184 

 465 



From this it would appear that the conductivity of the rocks 

 associated with the mineral mines is twice as great as that of the 

 artesian wells class. But all the experiments, with the exception 

 of three or four, were made with blocks of dried rocks, and those 

 showed a very remarkable difference ; thus, for example, dry 

 New Red Sandstone gave £o'00250, whereas when wet it was 

 increased to .£000600. The author remarks that as all rocks 

 below the level of the sea and that of the river valleys are per- 

 manently saturated with water, dry rocks are the exception and 

 wet rocks the rule in nature, consequently the inequalities of 

 conductivity must tend to disappear. The power of conduction 

 is also greater along the planes of cleavage or lamination than 

 across them, and therefore the dip of the strata must also exer- 

 cise some influence on the conductivity of different rocks and 

 "massifs." With respect to the foliated and schistose rocks, 

 M. Jannettaz has shown that the axes of the thermic curve along 

 and across the planes of foliation and cleavage are in the follow- 

 ing proportions : — 



Gneiss of St. Gothard I : i'5o 



Schists of Col Voza I : 1 So 



Cambrian slates, Belgium I : 1 '9S 



This cause will locally affect the rock masses. 



Conclusion. — The author deduces from the three classes of 

 observations a general mean thermic gradient of 48 feet per 

 degree F., but he considers this only an approximation to the 

 true normal gradient, and that the readings of the coal-mines 

 and artesian-well experiments are, owing to the causes he enu- 

 merates, still too high. He also discusses the question whether 

 or not the gradient changes with the depth. His own reduction 

 of the observations gave no result, but he points out that in all 

 probability the circulation of water arising from the extreme 

 tension of its vapour is stayed at a certain depth ; while it has 

 been shown experimentally that the conductivity of iron dimin- 

 ishes rapidly as the temperature increases, and this; may possibly 

 in a different degree apply to rocks. If, therefore, there is any 

 change, these indications would be in favour of a more rapid 

 gradient. 



Taking all these conditions into consideration, the author 

 inquires whether a gradient of 45 feet per degree would not be 

 nearer the true normal than even the one of 48 feet obtained by 

 the observations. 



Linnean Society, February 19. — Prof. P. Martin Duncan, 

 F.R.S., Vice-President, in the chair. — The Rev. L. Martial 

 Klein was elected a Fellow. — Mr. Thiselton Dyer exhibited 

 and made remarks on specimens of the peculiar Chinese "square 

 bamboo" (Bambusa quadrangularis, Fenzi), and of articles 

 made from the so-called "hairy bamboo" (probably Dendro- 

 calamus latiflorus, Munro), sent from Wenchow to the Kew 

 Museum by Dr. Macgowan. — Mr. T. Christy afterwards drew 

 attention to silk fibres received from Auckland, New Zealand. — 

 An abstract of Part iii. of the Rev. A. Eaton's monograph on 

 the Mayflies (Ephemerida?) was read by the Secretary. In this, 

 the fourth series of group 2 of the genera are dealt with. Among 

 representatives of Section 9, Chen is distinguished by absence of 

 hind wings, Callibeetis by costal projection and cross-veinlets of 

 its broad obtuse hind wings, Baetis by small or absence of costal 

 projection and deficiency of cross veinlets, and Centroptilum by 

 extreme narrowness of hind wings and slenderness of costal 

 projection. The distinctive characteristics of sections 10 and 1 1 

 of the genera are also taken into consideration, and full descrip- 

 tions of many new species given. — Then followed notes on the 

 European and North American mosses of the genus Fissidens, 

 by Mr. W. Mitten. Referring to the more recent important 

 contributions of Dr. Braithwaite's British Moss-Flora, and 

 Messrs. Lesquereux and James's North American Mosses, and 

 taking into account definitions of older writers, such as Dillenius, 

 Hedwig, Swartz, and others, Mr. Mitten endeavours to arrange 

 the entire group afresh, partly in a tabular form, and afterwards 

 supplementing this by notes on the individual species. — A paper 

 was read by Prof. P. M. Duncan on the anatomy of the Ambu- 

 lacra of the recent Diadematidae. The author described the 

 arrangement of the compound plates of the genera of Diadema, 

 Echinolhrix, Centrostephanus, Atropyga, Micropyga, and Aspi- 

 dadiadema. The first three genera have triplets, consisting of 

 primaries, the adoral and aboral plates being low and broad, 

 and the second, or central plate, being a large primary. Nea, 

 the peristome theie is deformity of this typical arrangement 

 and in Echinothrix a demiplate may enter, but it is never the 

 second plate. In Astropyga the triplets are arranged so that the 



