January 22, 1920] 



NATURE 



547 



struction of necessity results in a much stiffer and 

 more stable stage. There is, in fact, a general lack 

 of stability going through nearly all parts of a micro- 

 scope. But it is significant that, even so long ago as 

 the beginning of last century, the instrument as then 

 designed had much greater attention paid to this 

 point. The microscope an illustration of which 1 

 show on the screen is, to my mind, an embodiment 

 of a principle that should receive attention. So soon 

 as English makers are in a position to consider the 

 production of an instrument of a special type, it is 

 my intention to have one made. In this the general 

 principle is that all the optical parts are carried on a 

 bar which is, in effect, an optical bench, and that this 

 is strutted in such a way as to give stiffness to the 

 instrument as a whole. The only effort that I am 

 aware of that has been made in this direction is in 

 the microscope designed by Dr. Rosenhain, par- 

 ticularly for metallography, but which is adaptable for 

 ordinary work. This instrument, to my mind, is such 

 an improvement on any other type of stand that I am 

 at a loss to understand why metallographers have not 

 more generally taken it up. It might appear that I 

 ain exaggerating the importance of stability in the 

 stand, but it should be realised that any lack of 

 centration in the optical parts, or of alignment 

 in the optic axes of these parts, results in more serious 

 deterioration of the resulting microscopic image than 

 any other single factor. The optical parts of a micro- 

 scope are the objective, for obtaining the primary 

 magnified image of the object ; the ocular, for further 

 enlarging that image and transmitting it to the eye ; 

 and the sub-stage condenser, for illuminating the 

 object with a larger or smaller cone of light. The 

 limitations of time will prevent me from dofng more 

 than refer very briefly to some properties of the optical 

 parts. 



It is generally assumed that magnification is the 

 primary function of an objective, but in point of fact 

 the main point is not magnification, but resolution. 

 By resolution is meant the power the objective has 

 of separating and forming correct images of fine 

 detail. The theory known as the Abbe diffraction 

 theory is the one on which modern optical calcula- 

 tions are based ; and it is safe to say that it was 

 never more fullv accepted than at the present time, 

 and never rested on a surer basis. There has been 

 much discussion in this country of that theory, and 

 probablv a good deal of misconception has arisen from 

 its inapt designation, for the term "diffraction 

 theorv " is perhaps somewhat unfortunate. I cannot 

 do better than quote the late Lord Rayleigh in refer- 

 ence to this matter. He said: "The special theory 

 initiated by Prof. Abbe is usually called the diffraction 

 theorv, a nomenclature against which it is necessary 

 to protest. Whatever may be the view taken, any 

 theorv of resolving power of optical instruments must 

 be a diffraction theory in a certain sense, so that the 

 name is not distinctive. Diffraction is more naturally 

 regarded as the obstacle to fine definition, and not, 

 as with some exponents of Prof. Abbe's theory, the 

 machinery by which good definition is brought about." 

 This very clearly and accurately sums up the position. 

 The Abbe theory tells us that there are two main_ 

 factors determining resolution ; that is, the numerical 

 aperture of the objective used and the wave-length of 

 the light. Numerical aperture is determined for us bv 

 the optician, and it is well known that, with an oil- 

 immersion objective, a numerical aperture of 1-4 is 

 at the present time the practical limit. Metallo- 

 graphers are in a somewhat stronger position, as a 

 monobromide of naphthalene immersion objective was, 

 and presumably still is, made by Zeiss which had a 

 numerical aperture of i-6. This represents the abso- 

 lute limit at the present time, and there is no indica- 



tion that numerical aperture will be increased in this 

 sense by present methods. 



The other factor governing resolution is the wave- 

 length of light, and in this connection it must be 

 borne in mind that to resolve a regularly marked 

 structure the distance l>et^J'een the markings must be 

 more than half a wave-length. Under ordinary condi- 

 tions of illumination we cannot go very far in the 

 direction of increased resolution unless we resort to 

 an illuminant such as a mercury vapour lamp which 

 is rich in blue and violet radiations. There is much 

 room for investigation in this direction, as the ideal 

 illuminant for microscopic work has yet to be found. 

 But I do not know of any one that approaches so 

 nearly to it as the one I have mentioned, the mercury 

 vapour lamp. It suffers only from one disadvantage 

 that I can see, and that is that the differentiation due 

 to staining is not so clearly brought out as when 

 ordinary light is used. But as staining is itself an 

 artificial process, and is simply done to differentiate 

 structures, it only means a certain amount of educa- 

 tion to enable us to appreciate the differences even 

 under the light from this lamp. The only stains 

 which it does not show quite well, or rather in which 

 the colour-tint is altered, are those in which red pre- 

 dominates. Any other colour is shown perfectly and 

 in proper gradation. The advantages of this illu- 

 minant are that it is even and uniform. It has a 

 fairlv large area, and can be used, therefore, for any 

 class of work. Its intensity can be varied within con- 

 siderable limits by having a resistance in series, so 

 that the current density is altered to suit the par- 

 ticular work under observation. Further, it is possible, 

 by interposing neutral screens, to vary the light inten- 

 sity if the electrical method is inconvenient. Owing 

 to its possessing practically no red radiations its mean 

 wave-length is shorter, and by using suitable screens 

 light which is truly monochromatic, yellow, green, 

 blue, or violet, can be obtained at will. These lamps 

 are made both in glass and quartz, but the quartz 

 ones are preferable, because they admit of the use of 

 heavier currents with greater luminosity; and, further, 

 thev have a much longer life. I have exhibited two 

 of these lamps, because I regard them as far in 

 advance of anv other form of light available to the 

 microscopist at the present time, whether he is a 

 biologist or a metallographer. 



The whole subject of illumination needs investiga- 

 tion also, because there is, I think, little doubt that 

 a modification in the intensity of the illumination of 

 any particular object enables us to use a larger light- 

 cone than we could do in ordinary circumstances 

 — that is, variation of the intensity is an alternative 

 to the use of the iris diaphragm in the sub-stage of 

 the microscope. But it is in the direction of using in- 

 visible radiations in the ultra-violet, or, it may be, 

 radiations which are still shorter than the ultra-violet, 

 that developments in microscopic work are, in my 

 opinion, likely to occur. 



There are two other points worth mention, which 

 I trust may be dealt with more fully in succeeding 

 papers. One is that, while the resolution limits are 

 so inflexible, that does not by any means apply to 

 mere visibility. By illuminating small particles by 

 means of an annular cone of rays — that is, what is 

 ordinarllv known as dark-ground illumination — or by 

 illuminating them at right angles to the optic axis of 

 the microscojie — what is known as the ultra-micro- 

 scopic method — particles of a very much smaller order 

 of size can be made visible. But we cannot tell any- 

 thing about their form, nor can we accurately tell 

 their size. We are onlv conscious of their mere 

 existence. 



Another point to remember is that magnification is 

 definitely limited to something like 750 diameters 



NO. 2621, VOL. 104] 



