NOTE: Unfortunately the photographs in Prof. Wilska's letter are not of adequate quality to reproduce from the photocopy I have available.

Nature (171, 353, 1953)

A New Method of Light Microscopy

Using annular apertures made in a thin layer of candle soot on a microscope objective lens and corresponding apertures in the condenser, an ordinary microscope can be made very suitable for the study of unstained specimens. There are two different procedures.

To make higher refracting objects appear dark against a light background, it is necessary to find the objective lens surface nearest to the image of the condenser aperture and soot it by short contacts with the flame, until the transparency has diminished by one-half. On a watchmaker's lathe, using a sharp tip of some soft material, a narrow ring-shaped area is freed from soot, the diameter of the ring being greater than half the exit pupil of the objective, that is at least twice as wide as the phase ring in conventional phase-contrast objectives. When compared with the latter, there are freedom from halo effects, increased  resolving power and a pleasing gradation for objects having different refractivity.

Images of greater contrast and with light values similar to those to which we are accustomed in the macroscopic world are achieved if a partly re­versed procedure is followed. The sooting is continued until the transparency has decreased to 8-9 per cent, after which the soot is wiped off, except for an area covering the image of the condenser ring aperture. As a result, the background of the microscope image appears in an agreeable sepia tone, often giving the illusion of sky-illuminated objects lying on the sand floor of a shallow sea. Many features which are generally invisible with the normal use of the microscope are rendered visible ; for example, influenza virus on red cell ghosts, bacterial flagella, sperm tail pencil, and delicate membranous margins in living thrombocytes (see photographs). Apparently the method can be adapted to ultra-violet microscopy and mirror optics as well.

The theoretical explanation of what occurs in each of the alternative procedures may not be an easy one, since the phase effects in microscopical observation are still incompletely understood¹. The arrangements described above are the result of some hundreds of trials, in which the conditions were varied one at a time.

A. WILSKA

Institute of Physiology,

University of Helsinki.

Nov. 17.

Nature (171, 697; 1953)

A New Method of Light Microscopy

The method described by Prof. A. Wilska' appears to be a special case of phase-contrast, generally known as ‘amplitude-contrast'. Although not usually employedd with fixed phase-contrast objectives, it ffrequently crops up in variable phase-contrast and interference systems. Oettlé, for example, used it in his experiments with a variable amplitude and phase microscope. The basic theory can be simply explained by a vector treatment similar to that given in detail elsewhere. Object details can be represented by points in or on the well-known vector circle, centre O, radius OM. In conventional microscopy, the origin remains at O and transparent phase-changing details such as P, which lie on the circumference of the circle, are all equidistant from O and appear with equal intensity, OP² (= OM²), so that contrast is zero.

The object of phase-contrast and interference microscopy is to shift O to some new position so that points on the circumference will no longer be equidistant from the origin, but will appear with intensities depending on φ, the phase change introduced by the detail. In ordinary phase-contrast this is achieved by rotating the vector OM, which represents the direct light, and changing its amplitude, that is, by altering the phase and intensity of the direct or diffracted components. We can, however, shift O by changing only the amplitude of OM without rotation. In this case pure amplitude-contrast is obtained and the new origin lies somewhere on OM. In Prof. Wilska's second method, the intensity of the direct light is reduced to about 1/12. The ampli­tude is thus 1√12 or 0.29. The new origin is O₂, where 0₂M = 0.29, OM being taken as unity. The background intensity is now O₂M², the detail intensity O₂P². Thus for all except very small phase changes, the image detail will be brighter than the background and it will not be possible to distinguish phase advances from phase retardations. If all the direct light were absorbed, the new origin would coincide with M, giving central dark-ground illumination in which theoretically all details will be bright on a black background.

In his first method, Prof. Wilska absorbs 50 per cent of the diffracted light. In this case the new origin will be at O₁, where O₁M = √2. This is an example of B-type amplitude contrast^2,4. The detail intensity O₁P² is now always less than the background intensity O₁M², so that dark contrast occurs. However, for small phase changes, O₁P is very nearly equal to O₁M, so that contrast is inherently poor, and becomes worse as the transmission of the diffracted light decreases, that is, as O₁ moves farther away from M. This latter method is therefore only useful for rather refractile objects such as thick cells and fibres.

The halo effect⁵ is inherent in all phase-contrast and interference systems in which there is incomplete separation of the interfering beams (in this case the direct and diffracted components). It becomes much more conspicuous with heavy absorption of the direct light and is particularly noticeable in Prof. Wilska's photographs, for example, around the sperm tail in Fig. (c) and the edge of the epithelial cell in Fig. (e). The halo will be much less obvious in B-type contrast as observed by Prof. Wilska, and it will also depend on the width of the annular ring in the objective, which determines the degree of overlap between direct and diffracted light.

These considerations are based on the assumption that it is possible to affect the transmission of the light wave without altering its phase. If the partly absorbing layer introduces a phase change, then some degree of phase-contrast will occur and O₁ and O₂ will no longer lie on OM. In Prof. Wilska's first method the phase of the direct light will be advanced, and in the second case it will be retarded, relative to the diffracted light. This will give rise to positive and negative phase-contrast respectively, thus heightening the effects already predicted.

Prof. Wilska's letter should be useful in directing the attention of microscopists to the value of amplitude-contrast and central dark-ground illumination in some cases. His photographs show that excellent results can be obtained by very simple means.

R. BARER

Department of Human Anatomy,

University Museum,

Oxford.

Feb. 21.

¹ Wilska, A.. Nature, 171, 353 (1953).

² Oettlé, A. G., J. Roy. Micr. Soc., 70, 232, 255 (1950).

³ Barer, in "Contrasts de Phase et Contrasts par Interférences" (Paris: Revue d'Optique, 1952); J. Roy. Micr. Soc., 72, 10, 81 (1952); and in the press.

⁴ Bennett, A. H., Osterberg, H., Jupnik, H., and Richards, O. W., "Phase Microscopy" (Wiley, New York, 1951).

⁵ Zernike, F., Physica, 9, 686, 974 (1942).

WITHOUT criticizing the vector treatment of amplitude and phase-contrast microscopy kindly given by Dr. Barer in his comment on my letter¹, I would like to make the following comments : (1) Using my second method, phase-advancing objects appear darker than the background. The soot layer appears to cause a considerable phase retardation in addition to the absorption. (2) If the absorbing annulus is made substantially darker, the resulting image becomes peculiarly glossy and too unsharp to be of any value. Thus the "central dark-ground illumination" does not seem to be very promising in this connexion. (3) According to Oettlé², as well as from my own observations, ordinary negative phase-contrast methods do not produce images comparable with those of positive phase-contrast in clarity. The inherent haziness and glare of the negative phase-contrast picture may be due entirely to the light reflected from the phase plate and converged back to the area under observation. Due to the reflectivity of absorbing annuli evaporated by ordinary means, this additional surface illumination may reach higher values than the light passing the objective annulus. My method does not suffer from this in the same degree since the reflectivity of soot is small.

A. WILSKA

Institute of Physiology,

University of Helsinki.

¹ Wilska, A., Nature, 171, 353 (1953).

² Oettlé, A. G:., J. Roy. Micr. Soc., 70, 232 (1050).