274 



nAtuke 



\yan. 17, 1889 



From these microscopic olaservations on the crystals of 

 the second period, we may conclude that they are formed 

 purely and simply by igneous action, without requiring 

 the hypothetical temperaturfes and pressures formerly con- 

 sidered necessary, and without that absolute repose re- 

 garded as needful for the regular crystallization of minerals. 

 We see, indeed, that the microlites are formed after the 

 outflow, at the normal barometric pressure and at a 

 temperature far from being as high as generally supposed, 

 and we witness the birth of the crystals during the very 

 flow of the lava stream. When the cooling is extremely 

 rapid, the microlites have no time to form, and the lava 

 cdn produce only crystallites. 



But the microscope enables us to determine the 

 chronology of the crystals in lava in a still more detailed 

 manner. We have already distinguished two great periods 

 in their history ; let us now indicate in a general way how 

 we may establish to some extent the date at which each 

 species of the two groups is separated from the magma. 

 Data leading to the determination of their relative age are 

 afforded by their inclusions. 



A crystal developed in a vitreous mass frequently 

 incloses particles of the medium in which it grows. In 

 this way certain sections under the microscope appear 

 penetrated with vitreous grains, imprisoned in the interior 

 of the crystals and frequently arranged along the zones 

 of successive growth. These inclusions prove that the 

 crystals in question were formed in a vitreous mass, 

 liquefied by heat. In other cases the inclusions are 

 mineral species in the form of microlites ; and it is clear 

 that they must have been anterior in date to the mineral 

 in which they are inclosed. Finally, in other cases a 

 species will mould itself around sharply defined crystals, 

 conforming to their outlines, and filling up all the spaces 

 between the minerals ; thus showing that the crystals are 

 of earlier origin than the surrounding mineral. 



From these facts, which speak for themselves, we have 

 been able to draw up chronological lists indicating the 

 relative date of crystallization of each species of the two 

 great periods. I will not stop to cite these lists, but we 

 shall soon see how the law which governs the successive 

 formation of the crystals, and their relative age, is evolved i 

 from synthetic experiments. i 



'I have traced in broad outline the history of a lava, 

 but have sketched only a few of the details which modern 

 researches on lithological phenomena have developed 

 w'x'Cix such startling reality : nevertheless, what we have 

 seen is sufficient to show in a striking manner tbe power 

 of analysis when supported by reasoning. I think I am 

 not wrong in saying that from this point of view the study 

 of a lava presents one of the finest examples of the applica- 

 tion of the inductive method to the natural sciences. We 

 hardly know whether to admire most the analytical pro- 

 cesses, or the subtlety of observation, or the logical 

 method by which the observed phenomena have been 

 brought into connection. 



Microscopic analysis, powerful as a method of investiga- 

 tion, has enabled us to trace with close exactitude the 

 progress of crystallization in a rock where the unaided 

 eye could discover only an indistinct and uniform mass ; 

 to penetrate into this marvellous tissue of volcanic pro- 

 ducts, where millions of polyhedra occur within the 

 volume of a cubic centimetre ; to determine with 

 matheinatical precision the nature of each of these 

 infinitely small bodies ; to track them to their birth, and 

 follow them throughout their development, tracing all the ! 

 modifications to which they have been subjected under i 

 the influence of physical and chemical agents. 



Nevertheless, to the conscientious and modest investi- 

 gator, how much still remains unknown in connection 

 with the history of volcanic products, though the field 

 seems so narrow, and has already been so well worked ! 

 What problems remain unsolved, even by the most re- 

 fined observation ! But when observation can no longer 



aid us, when we have exhausted all the resources of this 

 method of investigation, there yet remains the method of 

 synthetic experiment. This forms one step more on the 

 road which leads to i perfect knowledge of the pheno- 

 mena, and may conduct us to their definite solution. 

 But, in order that synthetic operations may attain this 

 end, they must be directed with dtie intelligence and 

 design. 



One of the essential conditions of a geological synthe- 

 sis, as Senarmont remarked, is, that each of the artificial 

 operations should be compatible with all the circumstances 

 traceable in the products of the natural operation. The 

 slags and scoria; of our furnaces, which, as we have 

 shown, are related to certain natural products, are, it is 

 true, the results of synthesis, but synthesis made at hap- 

 hazard ; and thus, notwithstanding their high scientific 

 interest, cannot be placed on the same level with the 

 synthesis of which I am about to speak, where the ex- 

 perimentalist, bearing steadily in view the problem which 

 he desires to solve, attempts to realize in the laboratory 

 the identical conditions which have surrounded the 

 formation of the natural products which he wishes to 

 imitate. 



In logical order, the synthetic methods follow the pro- 

 gress of observation and of analysis. But even in the 

 very infancy of geology there were certain powerfi^l minds 

 which foresaw, with the glance of genius, the part which 

 experiment was destined to play in that science. Buff'on 

 proved by experiment that granite and the principal 

 crystalline rocks are fusible, and that they were trans- 

 formed by fusion into a vitreous mass. Some years later, 

 Spallanzani performed an extensive series of experiments 

 on the fusion of lavas, in order to overcome the pre- 

 judices which prevailed respecting the cause of the heat 

 of eruptive matter. 



But it is especially to Sir James Hall that belongs the 

 honour of having, by his celebrated researches, intro- 

 duced experiment into geology. He demonstrated its 

 application in a masterly manner, and was led to sound 

 generalizations. We have here to notice in Hall's re- 

 searches only those which relate to the synthesis of rocks. 

 About the time when Spallanzani studied, by laboratory 

 methods, the conditions of the formation of lavas, the 

 illustrious Scottish geologist was busy fusing the eruptive 

 rocks in a vessel of graphite : he observed that the pro- 

 duct of this fusion, if cooled rapidly, became an amor- 

 phous vitreous mass, while, if cooled more slowly, crystals 

 were formed. James Hall had already observed by ex- 

 periment the capital fact for future synthesis that, in 

 order to regenerate the crystals of a rock which has been 

 fused, it is necessary to maintain the glass obtained by 

 the fusion at an elevated temperature, but yet a temper- 

 ature always inferior to that required for the fusion of 

 the rock. During this process, certain minerals crystal- 

 lize. These facts may be paralleled with the phenomena 

 which lavas display when their temperature is lowered 

 after their emission. 



Towards the commencement of this century, Gregory 

 Watt directed his researches in the same direction. He 

 experimented on masses of basalt, 700 pounds in weight : 

 these he fused, and allowed to cool during eight days 

 beneath a layer of charcoal, which was slowly consumed. 

 During this prolonged recuit, spherulitic concretions of 

 fibro-radiated texture, 6 centimetres in diameter, separ- 

 ated in the opaque black glass resulting from the fusion : 

 finally, the glass passed into a stony condition, assumed 

 a granular structure, and became charged with very thin, 

 crystalline lamella;. .At the same time, its magnetism 

 was increased, while its density rose from 2743 to 2-949. 



One conclusion from the researches of Watt, which 

 are closely related to those of Hall, is, that crystalliza- 

 tion may occur during the period when the fused matter 

 commences to solidify. 



At the time when the road to the synthesis of rocks 



