Soon the mathematicians were refining their designs, however, and more complex designs began to emerge. The Zeiss Planar of 1897, the Cook triplets, the Tessars, all utilized designs with either more glass elements or more air spaces, or both, resulting in more air to glass surfaces. The Planar had four groups, and therefore eight air-glass surfaces. The resulting flair was thus double that of the earlier cemented lenses.

While not, in the early days, fully understanding the physics of reflections, it was known nevertheless that each air to glass surface reflected four to five percent of the incident light. It was also known that certain films, such as the "bloom" of oxides and films due to the corrosive action of acids reduced the reflections. Those of you who have older classic lenses probably have seen the "tarnished" look of some lenses. Dennis Taylor of the Cook company experimented with and patented the induced oxidation of the lens surfaces in production in 1904, but the results were found to be unpredictable.

Because the benefits of multiple element lenses were just too much to ignore, the designers continued to design lenses with greater and greater complexity. In particular, the need for high aperture lenses for use on fixed-lens, small format and movie cameras led to many designs such as the Speed Panchro type, derivatives of the early Gauss lenses, and these lenses suffered seriously from reduced contrast.

Finally, in 1936, the Zeiss Company developed the modern method of lens coating, which involved evaporating magnesium fluoride in a vacuum and depositing a thin layer on the lens surface. By depositing a film of just the right thickness, the light reflected from the surface can be cut in half at a certain median wavelength. Wavelengths which are longer and shorter will benefit to a gradually lessor degree as they increase and decrease from this median wavelength. As I will discuss below, this median wavelength can be chosen by the designer according to his needs and the use to which the lens will be put.

To understand how this works, it is necessary to examine how reflections are formed at the glass surface. As light passes between any two media of differing indices, there will be a random secondary emanation of "light rays" in every direction on every point of the interface surface all of which occur at different times. Many, indeed most of the secondary rays on each side of the interface cancel each other out because of their equal but opposite forces. Those which do not cancel out are the resulting reflections which are seen at the surface of the interface. Only a finite number of rays, about 95% of primary rays which are transmitted, and about 5% of secondary rays which are reflected, survive this interference process.

If a layer of some material is deposited on the glass surface which has an index such that the reflections at the two interfaces, those of the primary substrate and those of the coating, will have about equal reflections, and if the thickness of the coating is just right, or one quarter of the wavelength of the color of light to be suppressed, rays reflecting from the substrate encounter the rays reflecting from the coating, and being out of phase, the forces can be made to cancel each other out. Rather than resulting in a reflected secondary ray, the primary, transmitted ray will be augmented instead, increasing transmission. Because one of the conditions for this system to work is the correct film thickness for any given wavelength, the suppression of the reflected secondary rays is most effective only with a certain film thickness and wavelength combination. This describes the original single layer coating. Remember that the index of all of the media, that of the air, the substrate and the coating are key to knowing just what degree of success will be achieved in suppressing the reflections. The greater the difference between any two media, the greater will be the resulting reflection problems to solve. It is an accident of nature that the highest index glasses, which uncoated would have the highest reflections, benefit most from coating with magnesium fluoride because magnesium fluoride has an index which is high enough to suppress reflections of these glasses. This also means that glasses with medium indices will not benefit to the same degree from a single layer coating.

In the days before multi-coating, when it was desired to balance the transmission of a lens, different thickness coatings would be given at each lens surface. A straw colored coating might be given one surface (transmitting more blue light), while a blue layer (transmitting more yellow-green light) might be given another surface. Thus, the total transmission of a lens could be made to be fairly neutral.

I would like to put some of this in perspective. During most of the history of lens coating, the coatings used were these single layer coatings. The multi layer coatings are a much more recent innovation, and will be discussed below. For the sake of comparison, let us consider the typical lens of the 1970s, when a new lens received typical coating of magnesium fluoride. A typical plasmat such as the Symmar has four groups, or eight surfaces. If, due the coating, each surfaces reflects about two percent per surface, the total losses due to reflections can be expected to be about 16%. Now let us compare this to the older style, fully cemented lens. In the Dagor and Protar, as mentioned above, there are only four air-to-glass surfaces. As an uncoated lens, such a lens might be expected to reflect about four percent per surface, or a total of 16% of total light reflected. In other words, a modern lens having twice as many air-glass surfaces, having received a single layer coating, can be expected to have roughly the same transmission and flair levels as an uncoated Dagor. This is why the Dagor and other similar lenses continue to be viable so many years after their original manufacture. A Dagor with a single layer coating will actually have less flair than a single coated modern plasmat. Recently, a major firm made a limited number of 14" Dagors which were multi-coated. This lens stands today as probably the highest contrast lens ever made (when used at f:16 or higher to eliminate residual spherical aberration), since multi-coatings reduce reflections to about one-half percent per surface, or a total of only 2 percent total losses in this lens.

Mention should be made here of the misconception many photographers have regarding coatings and color correction. As I have stated in earlier articles, color correction has nothing to do with coatings. Color correction is the bringing to a common focus, both longitudinally and transversely, of two or three colors of the visible spectrum (in achromats and apochromats, respectively). This is accomplished solely by the choice of suitable dispersions of the glass elements employed. While it is true that color work my be adversely effected by the use of uncoated lenses, this is entirely due to flair from the uncoated surfaces, not to the lack of color correction. In fact many photographers continue to use Dagor type lenses for color work, and find the flair levels to be acceptable.

Many photographers question the feasibility of having an older lens coated. Indeed, many lenses were retro coated during the fifties and sixties by firms such as Burke and James and others. One aspect of the typical single layer coating is just how it is applied. When the first coatings were applied, they were simply vacuum deposited, and the mechanical strength of such a coating was only fair. It was soon discovered that by baking such a coating, the film became denser and tougher. When lenses were retro coated, baking was often impractical due to cemented surfaces in the lens which would melt and become separated. Thus, the lenses would be "cold" coated, leaving a "soft" coating. Many of you have no doubt found lenses whose coatings have begun to degrade. Such coatings often have patches which resemble water marks where the coating has come off. In other cases, the coating has degraded generally, and no longer retains its optical qualities. In such cases, the lens seems to have a foggy appearance. The only remedy for this lens is to polish off the old coating. If you have purchased a lens which is suspected of having received a cold retro coating, take care not to use any lens cleaning fluid on it. Such agents usually contain mild alkali components like ammonia. The ammonia can attack the coating and ruin it.

As stated above, a cemented lens would have to first be de-cemented before receiving a proper, baked, hard coating. The labor of disassembling and then re-cementing such a lens is rarely worth the cost. Even if a lens has no cemented surfaces, it must first be removed from any metal hardware in order to be coated, since the lens must be heated and metal hardware in contact with the glass could cause the glass to crack during the heating process.

In a related subject, many of you have experienced problems with the formation of Newtons rings when film comes on contact with a glass surface. This often occurs when a a negative is placed in contact with the negative carrier of an enlarger or placed in a contact printer. As described above, if the medium between the glass and the negative (air) is just the right thickness, interference will occur, resulting in the colored rings which ultimately show on the final print. If the source of the reflections which give rise to the interference rings can be removed, the rings will not be produced. This cannot be done completely, of course, but if the glass is coated the rings will be greatly reduced. If the glass receives a single layer coating, the ring intensities will be cut in half. If the glass is multi coated, which reduces the reflections to about one half a percent (from 4%), the reflections, and therefore the Newtons rings will be reduced eight fold. We have tested this with several photographers, and they report no visible Newtons rings in their final prints. I personally use a heavy piece of single-layer-coated plate glass to make all of 8x10 contact prints with no visible rings.