tion on the dry-bulb thermometer, to bring it to what he terms its true mean, he proceeds to do the same to the mean result of the self-registering thermometer. As his faulty observations at Greenwich made it appear that the Six's self-registering thermometer always gave too high a mean temperature, in order to reduce that mean to strict conformity with his supposed true mean, as given by his two-hourly readings of the dry-bulb thermometer, he deducted from the self-registering mean a quantity equal to the difference between the mean readings of these two instruments. On the mean of the year that deduction amounts to 1°07; so let us for a moment see what effect Mr Glaisher's corrections (?) would have on our Scottish observations-whether they would cause them to agree more closely, which if they were really corrections they would do, or whether they render both mean temperatures unquestionably incorrect and quite diverse from one another. The following table, then, exhibits the mean results at Makerstoun and for all Scotland for the years previously quoted, with the corrections to each set of instruments required by Glaisher's Tables, and the asserted UNIFORMITY sought for by the application of these tables : observed, as Corrections for each by ... 45-05 44-93 47 80 47.5 46.70 460 48.00 47.9 46.60 4-65 -1.07 0-00-1-07 +06-107 +06-107+06-107+06 Results,........... 43.98 44 93 46-73 48:1 4563 46 6 46-93 48.5 45:53 47-1 By this table it is seen that, instead of Glaisher's corrections causing the observations made by the two kinds of instruments to agree, it causes them to differ most widely from one another. Yet the very object of applying these corrections at all was for the express purpose of making them correspond. This table, therefore, it is hoped, will satisfy even the most prejudiced, that to alter by Glaisher's Tables the means as observed, or, to use scientific language which mystifies while it prevents inquiry, "to apply a correction for diurnal range," is only propagating error. This table, above all, shows the error of altering the means of the self-registering thermometers, which were all correct before being altered by Glaisher's Tables, but are just as manifestly rendered incorrect when changed by these tables. Now, seeing this is undoubtedly the case in Scotland, and is proved to be so even with one of its most southern stations, Makerstoun, it is nearly certain that such will also be found to be the case with nearly all, if not all, the stations in England, including Greenwich itself. We are precluded, however, from demonstrating that such is absolutely the case, by the circumstance that since this mode of altering the mean results has been adopted, Mr Glaisher, both in the Greenwich Observations, and also in the Meteorological Reports and Tables for all England, has ceased publishing the observed means of all the different series of readings from the dry and wet bulb thermometers, so that he puts it out of the power of any one to test the applicability of his correctness. It ought to be laid down as a principle, that, if the mean temperature is to be estimated at all, it ought to be deduced from the series of observations made with one form of instrument alone-and the one least liable to error is the selfregistering thermometer, constructed according to Rutherford's, Phillip's, or Negretti's principle; observations made with Six's thermometer being utterly worthless. There is thus avoided all source of error from not being certain of the amount of diurnal range which exists at each station, or from neglect to read the instrument at the exact minute of time when it ought to be read-sources of error which always exist when the drybulb readings are made use of. This mode of estimating the mean temperature also renders easy and certain the comparison of temperatures at different stations, or different parts of the world; it also avoids all those errors which arise from blunders on the part of the calculator, who, if he makes use of tables, is apt to add a number when he ought to deduct it, or deduct it when he ought to add it; and who, if he used Glaisher's Tables, as originally published in the "Philosophical Transactions," would be sure unwittingly to commit mistakes, as these tables are themselves not free from serious blunders. What, however, I most especially desire to effect by directing attention to this subject is, that in all the Meteorological Reports which are published, the strict mean of each series of readings by the different thermometers-as observed— should be given without any correction whatever for supposed diurnal or monthly range. All such corrections may be false, as they may be deductions from observations not in themselves trustworthy, as was the case with those from which Glaisher's Tables were constructed. But when the original observations themselves are published, every one has it in his power at any subsequent period to correct them, should trustworthy tables of correction be drawn up; and, in the meantime, he can compare them with observations made with similar instruments in other situations-a thing put out of his power to do so long as the present practice is followed. It is to be hoped that the facts adduced have sufficed to prove that the mode of estimating the mean temperature employed by Mr Glaisher is quite inapplicable to Scotland, and only leads to erroneous results. These same facts also afford strong ground for concluding that the same will be found to be the case with the English observations. These facts seem further to prove that, in the present state of our knowledge, we cannot go far wrong in holding that to be the true mean temperature which is the strict mean of the self-registering maximum and minimum thermometric observations, when these instruments are of proper construction; and as this mode of estimating the mean temperature is simple, requires no correction whatever for diurnal range, is strictly comparable with similar observations made in other parts of the world, and is free from all those sources of error to which the dry-bulb thermometric observations are liable, it ought, in the meantime, to be adopted as the true mean temperature, and be quoted as such. 242 Remarks on the recent progress of Sanskrit Literature and Comparative Philology. By JOHN MUIR, D.C.L. I propose, in this paper, to give some account of the recent progress of Sanskrit literature, and of comparative philology. As, however, these subjects have not yet attracted in this country that degree of attention which their importance demands, it may be advisable to premise some information regarding the earlier stages of their cultivation. The existence of Sanskrit, as the sacred depository of Hindu literature, was known to the early Jesuit missionaries in India; and they had not only studied that language, but also, in one instance at least, made it the vehicle of conveying religious instruction. But the effective and continuous study of this important and venerable tongue may be considered to have commenced with Sir W. Jones, who landed in Calcutta in 1783, and was the translator of the Institutes of Manu," and of the drama of Sakuntala, and the author of various dissertations on the mythology of the Hindus, and the affinities of that people with other nations. This study was assiduously prosecuted by Mr H. T. Colebrooke, who (in addition to various other important services) was the first to supply a tolerably complete sketch of the character and contents of the Vedas (1805); and who afterwards expounded, with remarkable accuracy, the principles of the different systems of Hindu philosophy (1823-1827). The next scholar whom we must mention as having distinguished himself in the same field, is the present Boden Professor of Sanskrit at Oxford, Mr H. H. Wilson, who published in 1819 the first, and in 1832 the second edition of his "Sanskrit and English Dictionary;" in 1827 his "Select Specimens of the Theatre of the Hindus;" in 1837, Mr Colebrooke's translation of the "Sankhya Karika," or "Memorial Verses of the Sankhya Philosophy," with notes and illustrations by himself; in 1840, his translation of the "Vishnu Purâna," a system of Hindu mythology and tradition; and in 1850-1857, three volumes of his translation of the Rigveda. The only other British student of Sanskrit whom it is necessary to mention here is Dr Ballantyne, Principal of the Government College at Benares, who has within the last few years made some important contributions to our knowledge of Hindu philosophy. (See the last number of this Journal, p. 103). The study of Sanskrit was early taken up, and has long been pursued with vigour, on the continent. In 1808, F. Schlegel published his essay on the language and wisdom of the Indians. In 1823, A. W. von Schlegel published an edition and Latin version of the "Bhagavad Gita ;" and (1829-1838) two volumes of the "Ramayana," with a translation of the first. F. Bopp seems to have begun in 1816 the publication of his works on the grammar of the Sanskrit, and on comparative philology,a science towards the construction of which he has since then so largely contributed. The other foreign scholars who have laboured most assiduously and fruitfully in the cultivation of Sanskrit literature, are the late Frederick Rosen, editor and translator of the First Book of the Rig Veda (1838); the lamented E. Burnouf, editor and translator of part of the "Bhagavata Purana" (1840-1847), and the historian of Indian Buddhism; Lassen, the author of the "Indian Antiquities" (Indische Alterthumskunde, 1847-1858), an attempt to construct a continuous history of ancient India from the mass of heterogeneous data afforded by Indian literature, coins, rock and pillar inscriptions, &c.; Boehtlingk, editor of the Indian Grammar of Panini (1839, 1840), and (in company with Roth) compiler of a Sanskrit and German Lexicon; Roth, the author of three dissertations on the Veda, and editor of the Atharva Veda, 1855-1856; Benfey, editor and translator of the Sama Veda; Max Müller, editor of the Rig Veda (1849-1856), and author of the "History of Ancient Sanskrit Literature" (1859); Weber, editor of the White Yajur-Veda, and its appendages (1852-1859), and author of a short history of Indian literature (1852); Goldstücker, the compiler of a new Sanskrit-English Dictionary on the basis of Wilson's; Gorresio, editor of the "Ramayana," with an Italian version (1843-1858); Dr E. Roer, who has rendered good service by editing various Sanskrit works for the Asiatic Society of Bengal, and by publishing some translations and dissertations within the last few years; and, finally, M. Adolphe Regnier of Paris, the translator of the "Prātisakhya Sutras." I shall now give some account of the most important labours of these scholars whose names have just been mentioned. A new era in the department of Vedic literature may be |