MICROSCOPE 265 tives both in England and elsewhere, with complete success. This method consists in the replacement of the water previously interposed between the covering-glass and the front glass of the objective by a liquid having the same refractive and dispersive powers as crown-glass, so that the rays issuing at any angle from the upper plane surface of the covering-glass shall enter the plane front of the objec tive, without any deflexion from their straight course, and without any sensible loss by reflexion, even the most oblique rays that proceed from the object keeping their direction unchanged until they meet the back or convex surface of the front lens of the objective. It is obvious that all the advantages derivable from the system of water- immersion will be still more thoroughly attained by this system of " homogeneous " immersion, provided that a fluid can be obtained which meets its requirements. After a long course of experiments, Professor Abbe found that oil of cedar wood so nearly corresponds with crown-glass, alike in refractive and in dispersive power, as to serve the purpose extremely well, except when it is desired to take special advantage of the most divergent or marginal rays, oil of fennel being then preferable. There are, however, strong objections to the use of these essential oils in the ordinary work of research ; and it seems not unlikely that a solution of some one or more saline substances will be found more suitable. In addition to the benefit conferred by the water-immersion system, and more completely attained with the homogeneous, it may be specially pointed out that, as no correction for the thickness of the covering- glass is here required, the microscopist can feel assured that he has such a view of his object as only the most per fect correction of an air-objective can afford. This is a matter of no small importance, for while, in looking at a known object, the practised microscopist can so adjust his air-objective to the thickness of its covering-glass as to bring out its best performance, he cannot be sure, in regard to an unknown object, what appearances it ought to pre sent, and may be led by imperfect cover-correction to an erroneous conception of its structure. It has been recently argued that, as the slightest variation in the refractive index of either the immersion fluid or the covering-glass, a change of eye-pieces, or the least alteration in the length of the body in a word, any circumstances differing in the slightest degree from those under which the objective was corrected must affect the performance of homogeneous-immersion objectives of the highest class, they should still be made adjustable. The truth of this contention can, no doubt, be proved, not only theoretically, but practically, the introduction of the adjustment enabling an experi enced manipulator to attain the highest degree of perfection in the exhibition of many mounted objects, which cannot be so well shown with objectives in fixed settings. But it may well be questioned whether it is likely to do the same service in the hands of an ordi nary working histologist, and whether the scientific investigator will not find it preferable, when using these objectives, to accept what their maker has fixed as their point of best performance. The principal source of error in his employment of them lies in the thickness of the optical section of the object; for the rays proceeding from its deeper plane, having to pass through a medium intervening between that plane and the cover-glass, whose refractive and dispersive indices differ from those of the glass and immersion-fluid, cannot be brought to so accurate a focus as those proceeding from the plane immediately beneath the cover-glass. The remedy for this, how ever, seems to lie rather in making the preparation as thin as possible than in the introduction of what is likely, in any but the most skilful and experienced hands, to prove a new source of error. Every one who has examined muscular fibre, for example, under a dry objective of very high power and large aperture, well knows that so great an alteration is produced in its aspect by the slightest change in either the focal adjustment or the cover-correction that it is impossible to say with certainty what are the appear ances which give the most correct optical expression of its structure. This being a matter of judgment on the part of each observer, it seems obvious that the nearest approach to a correct view will be probably given by the focal adjustment of the best homogeneous immersion-objectives, in fixed settings, to the plane of the prepara tion immediately beneath the cover-glass (see Jour. Roy. Micros. Soc., 1882, pp. 407, 854, 906). In every particular in which the water-immersion system is superior to the dry, it is itself surpassed by the oil or other homogeneous system, the anticipa tions of those by whom it was suggested being thus fully realized. But the advantages already spoken of as deriv able from the use of the " immersion system " are altogether surpassed by that which the theoretical studies of Professor Abbe have led him to assign to it, and of which he ias practically demonstrated its possession. For he has shown (as will be explained below) that the interposition of either water or oil so greatly increases the real "aperture" of the objective that immersion-objectives may be constructed having a far greater virtual aperture than even the theo retical maximum (180) of the angular aperture of an air-objective. The same eminent physicist, working on the basis supplied by the mathematical investigations of Professor Helmholtz and himself on the undulatory theory of light, has further established an entirely new doctrine in regard to the production of highly magnified representations of closely approximated markings. All that has hitherto been said of the formation of images by the compound microscope relates to such as are produced, in accordance with the laws of refraction, by the alteration in direction which the light-rays undergo in their passage through the lenses interposed between the object and the eye. These dioptric images, when formed by lenses free from spherical and chromatic aberration, are geometrically correct pictures, truly representing the appearances which the objects them selves would present were they enlarged to the same scale and viewed under similar illumination. And we seem justified, therefore, in drawing from such microscopic images the same conclusions in regard to the objects they picture as we should draw from the direct vision of actual objects having the same dimensions. The principal source of error in such interpretations arises out of the "interference " to which the rays of light are subjected along the edges of the minute objects through which they pass, or along any such lines or margins in their inner part as are sufficiently opaque to throw a definite shadow. For every such shadow must be bordered, more or less obviously, by inter ference- or diffraction-spectra ; and thus the images of strongly-lined objects with very transparent intermediate spaces may be so troubled or confused by these " diffraction- spectra " as to render it very doubtful what interpretation is to be put upon their appearances. A good example of this kind is afforded by the scales of the gnat or mosquito, which are composed of a very delicate double membrane, strengthened by longitudinal ribs on both sides, those of the opposite sides uniting at the broad end of the scale, where they generally terminate as bristle-shaped appendages beyond the intermediate membrane. These are crossed by fine markings, which are probably ridge-like corrugations of the membrane, common to both sides of the scale. Between each pair of longitudinal ridges there may be seen, under certain adjustments of focus and illumi nation, three uniform parallel rows of beads, which have been supposed to represent a true structure in the membrane. By Dr Woodward (colonel in the United States army), however, it lias been shown that this beaded appearance is merely the result of the "inter ferences " produced by the longitudinal and transverse lines of the scale. For the longitudinal diffraction-lines are clearly seen, alike in the microscopic image and in photographs (fig. 13), to extend into empty space beyond the contour of the scales, almost as far as the ends of the bristles in which the parallel ribs terminate; and they vary in number with the varying obliquity of illumination, so that in the same scale two, three, four, or even five rows of beads can be seen, and photographed at pleasure, in every intercostal space. 1 Every microscopist who has worked much with high powers is well aware of the difficulty of distinguishing between real and spectral markings, a difficulty which can only be overcome by training and experience. It seems, 1 Monthly Micros. Jour., vol. iv. (1876), p. 253.
XVI. - 34