pieces. The author's aim in its construction was solely the reinforcement of the impression by presenting an image to each eye, for he assumes, according to the then prevalent error, that vision by the two organs conjointly is naturally and necessarily unique, from the perfect conformity of all the homonymous parts of the two images of the ob ject on the two retina. The real advantage of such an instrument entirely escaped his attention; viz., that of presenting to the two eyes the two dissimilar microscopic images of an object, under precisely the same circumstances as the two unlike images of any usual object is presented to them when no instrument is employed, by which simultaneous presentment the same accurate judgment as to its real solid form, and the relative distances of all its points, can be as readily determined in the former case as in the latter.

In the construction of a binocular microscope there is one thing especially to be attended to-viz., that the images be both direct, for in this case only a true stereoscopic representation will be obtained. If the images, on the contrary, be inverted, a pseudoscopic effect would be produced which will give a very erroneous idea of the real form. The reason of these effects is fully explained in Sections 5, 10, 22, 23, of my Memoirs. The reversal of the images by reflection from mirrors or reflecting prisms, will produce the same result as to the stere oscopic and pseudoscopic appearances as their inversion by lenses. The binocular microscope constructed by the Père d'Orléans was pseudoscopic, though he describes one which, had it been made, would have been stereoscopic; he was, however, quite unaware that there would be any difference of this kind between them. The pseudoscopic effects when inverted images are presented, and the natural appear ances when erecting eye-pieces are employed, have not escaped the observation of Mr. Riddell.

Besides actual inspection by means of the binocular microscope, there is another way in which the advantages of binocular vision may be applied to microscopic objects. The beautiful specimens of photog raphy, reproducing the highly magnified images of objects, inserted in a recent number of the Microscopic Journal, makes one regret that they were not accompanied by their stereoscopic complements. A very simple modification of the usual microscope would fit it for producing the two pictures at the proper angles; all that is necessary is to cause the tube of the microscope to move independently of the fixed stand round an axis, the imaginary prolongation of which should pass through the object. A motion of 15° would include every difference of relief which it would be desirable to have, and it is indifferent in what direction this motion is made in respect to the stand. The pair of stereoscopic pictures may be obtained by a still simpler method, which requires no alteration in the microscope; the object itself may be turned round on an imaginary axis within itself, from 7° to 15°. But this method is inapplicable unless the light be perfectly diffused and uniform so as to avoid all shadows, the presence of which would give rise to false stereoscopic appearances. In the former case, where the object remains stationary and the tube moves independently of the frame, the arrangement of the light so as to cast single shadows might be an advantage, and assist the visual judgment.

5. On the periodic and non-periodic variations of Temperature at Toronto in Canada from 1841 to 1852 inclusive; by Colonel EDWARD SABINE, R.A., Treasurer and Vice-President of the Royal Society, (Phil. Mag., [4], v, 376.)--The principal object of this communication is to make known the non-periodic variations of temperature for every day in the twelve years, from 1841 to 1852 inclusive, at Toronto in Canada. The non-periodic variations are those differences of the temperature from its mean or normal state which remain after all the known periodical variations are allowed for, and are such as are generally accompanied by peculiarities of wind or of other meteorological circumstances. Recent investigations have led to the inference that opposite conditions of weather prevail simultaneously in the same parallels of latitude under different meridians, and that in particular Europe and America usually present such an opposition, so that a severe winter here corresponds to a mild one there, and vice versά; and recent theories of the distribution of heat on the surface of the globe profess to furnish the explanation. To place the facts on indisputable ground, it is requisite that a comparison should be made of unexceptionable records of the non-periodic variations in Europe and America, continued for a sufficient time to afford a proper basis for inductive generalisation. Toronto, from its latitude 43° 40′ N. and inland situation, is well suited to supply such a comparison with stations in the middle parts of Europe where similar records have been kept; and the twelve years embraced by the observations, viz.: 1841 to 1852, have been years of unusual meteorological activity in Europe.

Details are given in the commencement of the paper showing the care bestowed on the examination of the thermometer employed, with a standard divided" à l'échelle arbitraire," by the method of M. Regnault; as well as the precautions adopted for its fair exposure, and for its protection from rain and radiation. The observations were made by the non-commissioned officers of the detachment of the Royal Artillery employed in the duties of the observatory.

The period of twelve years comprises two series, in one of which the thermometer was observed hourly, and in the other less frequently, each observation in the second series receiving however a correction to the mean temperature of the day furnished for every hour and every day of the year by the first series. The two series, each of six years, are separately discussed; from the first series equations are derived. from the mean monthly temperatures by the method suggested by Bessel (Astron. Nach. No. 136), whereby the most probable values of the temperature, on every day and every hour, are computed corresponding to the whole body of the observations. These the author regards as approximate normal values, and by comparing with them the actual daily temperatures,-which in the first six years are the means on each day of twenty-four equidistant observations, and in the second six years the means of all the observations made on each day, each observation having been corrected for the hour in the manner described, the nonperiodic variations for every day in the year are obtained and are given in a table.

From the approximate normal temperatures the author has represented in a plate the phenomena of the temperature at Toronto, according

to a method which, if applied to the different meteorological elements and in different localities, might, he thinks, materially facilitate their intercomparison. This method, in which three variables are represented, one being dependent on the other two, is essentially the same that has been long used in magnetic maps, and in the ordinary isothermal maps; from which latter however it differs in this respect, that, whereas in the ordinary isothermal maps the two variables on which the variation of temperature is dependent are the geographical latitude and longitude, in the present case the two variables are the hour of the day and the day of the year. The variation of temperature is here referred therefore to time and not to space; a distinction which the author proposes to convey by employing the term Chrono-Isothermals, as applicable to lines of this description. From the delineation in the plate, and from the tables contained in paper, many characteristic and some peculiar features of the climate and meteorology of the part of the North American continent in which Toronto is situated, are readily perceivable. Several instances are pointed out; amongst these may be noticed the peculiar anomaly of the North American winter, which is very conspicuous in the plate; and the absolute as well as relative variability of the temperature at different seasons of the year, exhibited by means of a numerical index analogous to the probable error of the arithmeti. cal mean of a number of partial results, and deduced in a similar manner from the differences of individual years, months, and days, from their mean values: whence it appears, in respect to the annual temper ature, for example, that in any particular year there is an equal proba bility that its mean temperature will fall within the limits of 43°8 and 44° 6, as that it will exceed those limits on either side.

Finally, the author has shown the "Thermic Anomaly" (as it has been recently termed) of the monthly and annual temperatures at Toronto by comparison with the normal temperatures computed by Dove (Verbreitung der Wärme, 1852), for the parallel of 43° 40' N. from 36 equidistant points on the parallel; from which comparison it appears that after allowance has been made for the elevation above the sea (342 feet), every month of the year is colder than the normal temperature of the same month in the same parallel; that the thermic anomaly reaches its extreme in February, when it exceeds 10° of Fahrenheit; and that on the average of the whole year it is little less than 6°.

6. On Periodical Laws in the larger Magnetic Disturbances; by Captain YOUNGHUSBAND, R.A., F.R.S., (Phil. Mag., [4], v, 379).—In this communication the author has arranged, in tables, the disturbances of the magnetic declination at St. Helena and the Cape of Good Hope, for the purpose of exhibiting the systematic laws by which those phenomena are regulated, which were long described as irregular variations, because they were of occasional and apparently uncertain oc

currence.

The frequency of the disturbances, and their amount, whether viewed separately as easterly or westerly movements, or as general abnormal variations (easterly and westerly being taken together), is shown to be dependent upon the hour of the day, the period of the year, and upon the year of observation. This dependence upon the year of observation affords additional testimony of a periodical variation in the magni

tude of magnetic changes of the same character as that which has been found to exist at other places, and which has been considered to be coincident with variations of the solar spots.

The disturbances of larger amount only are noticed; those observations which differed by 2.5 scale divisions (18 in arc at St. Helena, and 1'9 in arc at the Cape) and upwards, from the normal place, were separated from the others and the values of the differences taken; there were therefore two series of figures to be dealt with, viz: the number of disturbances, and the aggregate amount of disturbance. These were separated into disturbances of the north end of the magnet towards the east and towards the west, and the effect of each considered separately.

The periodical character of disturbances at St. Helena and the Cape in a cycle of years is indicated so far as the limited extent of the observations would permit; sufficient however to point to the year 1843 as that of least disturbance at these two places, by showing a regular decrease from the previous years, and an increase in every succeeding year of observation. Though the hourly observations were discontinued before 1848, the year which Colonel Sabine has shown to be that of periodical maximum, (as 1843 was that of minimum magnetic activity at Toronto and Hobarton,) the observations now discussed are shown to be quite consistent with this period, and thus tend to establish it as a general law of magnetic phenomena. In the aggregate of each year the disturbances towards the west are shown to preponderate over those towards the east, both at St. Helena and the Cape of Good Hope; a similar preponderance of westerly over easterly has been found in every year of observation at Hobarton, but at Toronto the easterly distur bances exceeded the westerly both in number and amount in every year.

Arranging the disturbances into the several months of their occur. rence, the greatest disturbance is found to occur in January and the least in June at St. Helena and the Cape of Good Hope; the same months being those of greatest and least disturbance at Hobarton, whereas at Toronto, both January and June are months of minimum disturbance, the maxima disturbance occurring there in April and September.

From this identity of the epoch of greatest and least disturbance,at St. Helena, where the months of January and June are not those of opposite seasons, viewed either with respect to the sun's extreme altitude or to extreme periods of temperature,-at the Cape, situated in S. latitude 33° 56',--and at Hobarton in S. latitude 42° 52',-and contras ting this identity with a different law at Toronto in N. latitude 43° 39′, the author infers that the principal causes which produce an annual period of disturbance are not dependent upon local seasons. It is likewise pointed out that about the period of the equinoxes there is a tendency to maximum disturbances at all the stations, producing absolute maxima at Toronto, faintly but systematically indicated at the other stations.

The westerly disturbances were found to exceed the easterly in every month in the year at St. Helena and the Cape, which agrees with the results deduced from the Hobarton observations, while it appears SECOND SERIES, Vol. XVII, No. 49.-Jan., 1854.

19

from the observations at Toronto that the easterly disturbances exceeded the westerly in every month. The average value of a westerly distur bance is greater than that of an easterly in every month at St. Helena and the Cape of Good Hope. The disturbances at Hobarton again coincide with this result; and in a slight and less perfectly marked degree, Toronto has the same peculiarity.

Arranging the disturbances into the several hours of their occurrence, the hours of the day are found to be those of greatest disturbance in a very considerable degree; the sum of the ratios, during the twelve hours of the day, being about seven times as great as the sum of those in the twelve hours of the night at St. Helena, and about 2.6 times as great at the Cape of Good Hope; while at Hobarton the sum of the twelve night ratios slightly exceeded the day; at Toronto the excess was larger, viz: as 13 to 1. The laws of easterly and westerly disturbances, in relation to the local hours, are then examined separately. At St. Helena and the Cape, the easterly day disturbances exceed the easterly night-disturbances, and the westerly day-disturbances exceed the westerly night-disturbances. These results are compared with those at Toronto and Hobarton.

At St. Helena, although but comparatively few disturbances occur during the night hours, those disturbances are almost all westerly (183 disturbances, in all, occurred in nine night hours during five years, of which 174 were westerly and but nine easterly). In the day hours the westerly only slightly exceed the easterly disturbances. At the Cape, the westerly excess is less in the night and greater in the day than at St. Helena, and the night excess much greater than the day excess.

At St. Helena, the fact of the disturbances being more frequent in the day than in the night is consistent in every month of the year; this appears worthy of remark when it is remembered that at St. Helena the curve of the diurnal variation of the declination is precisely reversed at two opposite periods of the year; in one case corresponding to the curve of diurnal variation in middle northern latitudes, and in the other to that in middle southern latitudes. .

The mean effect of the disturbances which have been separated as described, and which comprise all of largest magnitude, is a constant westerly effect at every hour both at St. Helena and the Cape of Good Hope, acting more energetically in the night than in the day. At Toronto, the mean effect is westerly in the day and easterly in the night; at Hobarton, easterly in the day and westerly in the night.

7 Mode of Determining the Optical Power of a Microscope; by Prof. HARTING, (Quart. J. Mic. Sci., July, 1853, 292.)—l conclude by noticing another method of testing the optical power of the instrument, which, although rather troublesome, appears to me among the best, permitting us, as it does, to ascertain with a great degree of accuracy and certainty, the utmost limits of penetrating and separating pow er possessed by a microscope, and hence easily to express numerically its optical qualities in the most varied circumstances.

This method consists simply in subjecting to observation under the microscope the dioptric images of certain minute objects instead of the objects themselves. These images can be diminished at pleasure by withdrawing to a distance from the lens the object which forms them;