The remark, that the observations at Greenwich are commonly concluded at midnight, would be of some weight, if it could be proved that any thing essential is omitted by this practice, which does not appear to me to be the case. The observations relate chiefly to the sun, the fundamental stars, the moon, and the oppositions of the planets; and it may easily be discovered that these different series are exhibited with an uncommon degree of perfection. Had the censor in the Philosophical Magazine pointed out any other series of observations which could have been combined with these, so as not to interfere with them, no doubt the Astronomer Royal would have been much obliged to him. Every thing cannot be done at once in an observatory; and if as much is effected as can be wished in one respect, something must be omitted in others. But to multiply observations, without any plan or object whatever, would be mere idleness. Whoever is dissatisfied with the actual riches of the Greenwich observations, would do well to make the attempt to excel them; he would convince himself by such an experiment that the labour and patience required for doing so much, are fully sufficient to exhaust the powers of any one man. The third class of errors, relating to the meteorological instruments, I have not yet mentioned, because I think myself that greater accuracy is required in this department than it has hitherto been usual to observe. And if I should be allowed to suggest any improvement that could be made in the observations at Greenwich, it would be a more correct account of the meteorological instruments, and of the place in which the exterior thermometer is fixed. [It may, indeed, be expected with confidence that Professor Bessel's desire to possess a barometer and a thermometer, correctly compared with those which are employed at Greenwich, will not long be allowed to remain ungratified, though it would be a subject of much surprise on this side of the Channel, if he should detect in them such discordances as he is inclined to suspect.]

[This letter has probably appeared in Professor Schumacher's Nachrichten, though the 84th number of that interesting collection, for which it was intended, has not yet reached this country.]

113

ART. XIV. ANALYSIS OF SCIENTIFIC BOOKS.

I. An Attempt to establish the First Principles of Chemistry by Experiment. By Thomas Thomson, M.D. Regius Professor of Chemistry in the University of Glasgow, F.R.S., Lond. and Edin., &c. &c. In Two Volumes. London, 1825.

THE well-known author of this work regards the soul and body of chemistry to consist in a knowledge of the relative weights of the combining substances. This is to form a very narrow conception of the science. The true function of the chemical teacher is announced in the following verse of the Roman poet :

In nova fert animus mutatas dicere formas

Corpora.

It is the characteristic of chemical genius to reveal new elementary bodies, to form new compounds of the elements known before, to discover new qualities and relations both among simple and complex substances, and to arrange the manifold and marvellous phenomena of corpuscular action, under a few general laws. The philosopher of ardent and inventive mind, content to know the general proportions, is unwilling to stop his career of discovery in order to learn the minute fractional quantities; nor will he suffer his whole faculties to flutter round the oscillations of a balance. Let none, however, hence imagine, that we desire to disparage quantitative research; we would only assign it a place of due subordination below the qualitative, conversant with new powers and forms of matter. To view, with Dr. Thomson, the first principles of chemistry as consisting in an enumeration of weights and measures, is to narrow and debase the science into an affair of addition and subtraction. This arithmetical process is, no doubt, a valuable accessory; but can never compete either in interest or utility with the knowledge of the chemical affinities, from whose play, the countless diversities of composition and unceasing successions of form in the material system, are derived. These are the grand principles of chemical action, an acquaintance with which must necessarily take precedence of the study of quantity.

It is deeply to be lamented that the latter kind of inquiry, which, as exhibited in the work before us, can hardly be deemed an intellectual operation, should have usurped, to too great a degree, in some recent publications, the place of researches into the powers that modify matter. Admirable specimens of this sublime study are to be found in the statics of M. Berthollet, the Bakerian lectures and "Elements" of Sir H. Davy, and in many memoirs VOL. XX.

I

of M. Gay Lussac and Dr. Wollaston; and the individual, who should collect and arrange these into a compendious volume, would do no mean service to science. Of such philosophical principles we can perceive no trace in Dr. Thomson's book; though the modern multiplication of chemical objects, simple and compound, loudly demands their general connexions and dependencies to be developed.

"An Attempt to establish the First Principles of Chemistry by Experiment," is the title of Dr. Thomson's book; a title which may be understood in two senses. First, it may seem to imply that the doctor modestly offers his labours as a mere attempt; or secondly, that others before him might perchance have laid down first principles of chemistry, but that the grand æra of establishing them by experiment, was reserved for himself. The reader is not left long in suspense by the ambiguity of the title; for the first pages show that the latter conceit has taken possession of the author's mind. From the preface, indeed, one might be led to look for some of those Herculean achievements with which Sir H. Davy astonished the world in his Elements of Chemistry, although his title page did not blazon them forth. But nulla fides fronti is an adage which the reviewer has too many reasons to recollect.

We shall not, however, deal ungenerously with the Doctor; far less mete out to him with his own measure. We are ready to admit that by noting the mutually precipitating quantities of two neutral salts, he has in several cases given useful corrections of the primitive combining weights of bodies; and that he has, on some occasions, shewn errors in the second, and even first, decimal places of numbers formerly found. But, undoubtedly, his chief ingenuity is displayed in concealing, throughout the details of the book, the previous researches on the same topics of other experimenters, even when their results do not perceptibly differ from his own, which he presents as absolute perfection. Hence a young student will be led by the perusal of Dr. Thomson's "Attempt," to consider it both as the commencement and completion of che mistry, since he deigns to notice few precursors, and those chiefly for the purpose of pointing out their mistakes. He demonstrates his own numbers to be true, frequently to a millionth part; and rarely rests satisfied with a possible error of one part in a thousand!

Nothing places in so strong a light the vast improvement of practical chemistry during the last thirty-five years, as a compa rison of modern results on chemical equivalents, with those obtained by Richter in his Researches published in 1792, and some succeeding years. He mixed together two neutro-saline solutions, in such a proportion, as to produce their mutual decomposition, as indicated by the complete precipitation of the new-created com

pound. This chemical operation is very simple, and with regard to many substances, susceptible of very considerable precision. Yet Richter's equivalent numbers are so erroneous, as to shew that he must either have been very careless, or have employed very impure salts. Wenzel, a much earlier writer, and the real author of those general views which Richter prosecuted concerning mixed saline solutions retaining the state of neutrality or acidity which they previously possessed, had however made far more accurate researches on the composition of salts; but these had been unaccountably neglected, though capable of furnishing excellent data for the theory of chemical equivalents.

Fischer, in the compendious table which he constructed from Richter's voluminous experimental tables, states sulphuric acid at 1000. Hence if we divide his equivalent numbers by 20, the quotients will shew their relation to sulphuric acid reckoned 50, as it is on Dr. Wollaston's scale. The following are a few of

39.6

Lime Surely nothing but the impurity of the bodies submitted to experiment, can account for the errors in these numbers; the three acids presenting the only tolerable approximations to truth. And indeed till chemicals could be procured in a state of purity, the method of research, by ascertaining the mutually precipitating quantities of saline matter, was quite nugatory. At the present day, however, the greater part of the most interesting saline compounds are prepared for sale by the manufacturer, so beautifully crystallized as to be quite free from impurities, and admirably adapted for the investigation of equivalent numbers. Such articles, made on the great scale by eminent dealers, are generally to be preferred for scientific purposes to those made in little capsules by the closet experimenter.

When the happy idea of atomic combination was broached by Mr. Dalton, chemical synthesis and analysis had become much more exact, as his collation of results exhibits. In the first volume of his "New System," published in 1808, we find the following numbers, reduced to oxygen 10, and sulphuric acid 50.

Azote
Carbon
Phosphorus

7.1 7.1

13.0

Magnesia
Lime

Soda

Potash

Nitric acid

28.6

33.0

40

60

27.1

The only articles here very erroneously given are azote, and its compound, nitric acid. Most of the other numbers do not differ

much from the latest determinations; while those for soda and potash are exact.

Berzelius, familiar at an early period with Richter's speculations, was naturally prepared to embrace Mr. Dalton's views of atomic combination. He zealously set to work to determine, by the most refined and rigid methods of analysis, and synthesis, the true proportions in which chemical bodies combine. His successful labours formed the ground-work of Dr. Wollaston's valuable memoir on chemical equivalents. It is meanwhile worthy of remark, that Berzelius should very rarely have had recourse to Richter's method of mutual precipitation, in order to infer the atomic number of an unascertained salt, or of its constituents, from that of one already known.

Dr. Thomson is the first chemist who has methodically pursued this very simple and obvious route. Operating with the purer salts of modern times, he has been enabled to rectify and define the atomic numbers of a good many compounds; and thence also to deduce, on some occasions, the atomic weights of their constituents. He commenced this task in the "Annals of Philosophy, for Nov. 1820," in which he published the atomic weights of barytes, potash, soda, muriatic acid, protoxide of lead, sulphuric, nitric, and chromic acids. Of the numbers for the first four, as stated by Dr. Wollaston, he corrected the second decimal figure; the others seem to have admitted of scarcely any alteration, for Dr. Thomson's numbers nearly coincided with the previous determinations of Wollaston and Berzelius.

In verifying atomic numbers by the method of mutual saline precipitation, had Dr. Thomson been careful to ascertain that the decomposition was complete, as he might have done by suitable re-agents, much more confidence might have been reposed in his labours. But he has very frequently neglected this essential precaution, as we shall shew in the course of this examination of his work, and hence he has pretty often presented us with results, tallying well with the atomic theory, and with Berzelius, which he states as his own, though they could never have been derived from his narrated experiments. He has been also somewhat imprudent in quitting the solid track of precipitation, and of trying by novel methods, to demonstrate the truth of the idea suggested by Dr. Prout, that the atomic weights of all chemical bodies are multiples, by a whole number, of the atomic weight, of hydrogen. Ceratis ope Dædalea Nititur pennis.

The ingenious author of that proposition adduced so many examples, and analogies, as to render it highly interesting and plausible; and it has accordingly been regarded with a partial eye by the most eminent of our chemical philosophers. The beautiful law of gaseous combinations, discovered by M. Gay Lussac, and their