In my researches on the rediscovery of the old Italian violin varnish, it was essential to use only materials that were available during the years 1550 to 1750 when Stradivari, the Amati, the Guarnerei, etc. produced their masterpieces. For example, clay, wood ashes and madder root were the primitive sources for some of their raw materials. (1) But such sources are not satisfactory in modern times. One violin maker wrote that he tried to prepare an extract from clay and placed it to dry on the roof of his house.
He reported his results in this manner: “And then the rains came.”
It may be helpful to summarize the progress and the improvements that have been made after eleven years further research since my book was published and to describe modern adaptions so that violin makers can use them more readily. The scientific and chemical background should be explained. For the sake of brevity, it will be necessary in the following formulas and discussions to refer to portions of my book. (2) The original principles concerning the composition and the preparation of the varnishes remain valid and have been confirmed by analyses of old Italian varnishes. (3,4,5) Linseed oil and the metal rosinates remain as the fundamental ingredients; however, boiled oil is now preferred because of its better wearing quality and because clear films are obtainable with the resins to be described in this article. (See pages 56 and 57). The oil may be boiled over a small amount of umber in which case ageing for clarification is necessary. The ratio of oil to resin remains three to two.
Some changes have been made in the preparation of the metal rosinates which impart improved properties to the varnishes.
(1) Preparation of Rosin Solution.
275 cc’s Water (distilled preferred).
13.8 grams Potassium Carbonate (c.p. K2CO3)
30.2 grams Gum Rosin (ww grade)
300 cc’s Cold Water for chilling
Dissolve the potassium carbonate and the rosin in 300 cc of water by boiling in a liter Pyrex beaker. This requires a few minutes depending on the size of the lumps of rosin which should be clear, pea to nut size. Boil for five minutes more. Add 300 cc of chill water and make to 600 cc final volume.
The substitution of potassium carbonate for potassium hydroxide (Preparation 1, page 38) may not appear important at first glance. Potassium rosinate is formed in both instances. Moreover, the foregoing formula is very similar to Preparation 8 on page 49; it differs only in the amount of potassium carbonate; this gives the incidental advantage that the time for the rosin to dissolve is greatly shortened. The explanation for the change can be understood from the chemical reaction:
K2CO3       plus     Rosin
    138                     302
KHCO3 (Potassium Bicarbonate) plus Potassium Rosinate The formulation of the potassium bicarbonate is important in the precipitation of the resins which will be explained later.
(2) Preparation of Precipitants.
100 cc 5% Aluminum Chloride Solution (AICI3 H2O) 30 cc 5% Calcium Chloride Solution (CaCl2)
Five percent solutions are prepared by adding 5.0 grams of each salt separately to each 100 cc water. The foregoing mixture contains nearly equal amounts of aluminum and calcium.
Calcium chloride has been added to the precipitating solution because analyses of old Italian varnishes disclosed the prese4nce of calcium compounds, and further research has shown that their inclusion is desirable when used in conjunction with other metals; the use of calcium rosinate alone in the varnishes was not found satisfactory. (Preparation 31, page 84). The formation is of colored compounds from aluminum and calcium with alizarine is still not thoroughly understood (6) but the combination does yield lightfast colors.
Since calcium chloride is now present in the precipitant, alum, a sulfate, should not be used because of the formation of insoluble calcium sulfate upon standing. This is the only reason that alum, recommended in the book (pages 34, 39, 50 etc.) is now replaced by aluminum chloride.
(3) Precipitation of Resins.
100 cc’s potassium rosinate Solution (above) 60 cc’s Al-Ca Precipitant (above)
The two solutions are mixed whereupon a white precipitate is formed. This is treated according to Preparation 2, page 39. In fact, the precipitate contains a form of aluminum rosinate. The completeness of the reaction may be tested with blue litmus paper which should turn red.
When the precipitant is added to the potassium rosinate solution, normal metal rosinates are NOT formed. (See page 50). Instead, it has been found that basic metal rosinates result, perhaps A1(OH)(rosin)2 because of the presence of potassium bicarbonate. This is desirable because the resins are harder and the varnishes in which they are used are less affected by moisture. After the book was published, reports were received that varnishes, made from rosin solution according to Preparation 1, page 38, became soft and tacky in hot humid weather. This defect is eliminated by these modified resins.
Some violin makers question the use of rosin varnishes. This discussion should explain that rosin does not appear as free rosin in the final varnish resins but is combined with metals. The undesirable properties of rosin (high acidity, for instance) have been removed. Aluminum, iron and calcium rosinates are different from unreacted rosin for varnish making purposes.
Alizarine will be used to color the orange and red varnishes and the amounts to be used have been indicated in my book; the two percent suspension is still used and is added to the potassium rosinate solution in which it dissolves readily. Iron compounds remain as the coloring agent for the brown varnishes and the amounts of iron salts to impart various shares of brown have been outlined. However, iron chloride (FeHl3 6H2O) should now be used because ferric alum is also a sulfate and should not be mixed with calcium chloride in water solution.
The working life of the resins may be extended by storing them in a tightly stoppered bottle in a refrigerator. Turpentine remains as a satisfactory solvent for the varnishes in the proportions suggested in the book; however, it has been found that a solvent called dipentene imparts better “brush ability” to the varnishes because of its slower evaporation rate.
The varnishes, together with the preliminary treatment of the wood with raw linseed oil, produce a finish for violins, violas and cellos that is surpassingly beautiful and worth the time and effort to prepare it. And the violin maker who does use it will have the deep, heartwarming satisfaction that the finish is one that the old Italian masters might well have used.

(1) Michelman, Joseph. “Scientific Monthly”, Vol. 81, No. 5, November 1955, pages 221-223.
(2) Michelman, Joseph. “Violin Varnishes”, Cincinnati, Ohio. (1946).
(3) Michelman, Joseph. “Journal of the Franklin Institute”, June 1949, page 569.
(4) Letters, Karl. “Farbe and Lack”, July 1952, page 293.
(5) Michelman, Joseph. “Science”, September 22, 1950, page 337.
(6) Ciba Review, No. 39, May 1941, Page 1418 and others.

Making Michelman Violin Varnish (my way)
by Violinmaker Salve Haakedal         
(For those who can't get Joseph Michelmans book)

Making the base varnish:

60 cc Spirits of Turpentine
40 g  Mastic
Put it in a bottle and shake occasionally for a couple of days.

Add to it

60 cc Raw linseed oil
No need to filter: just let the gummy residue settle.

Making resin:

This resinmaking is the core of Michelmans varnish. It basically makes Aluminum Rosinate. This is a harder and more durable resin than pure Rosin. It also makes it possible for a color matter to dissolve in the varnish. The color from Madder attach to the Aluminum in the resin, forming a kind if Madder Lake.

First make 2 solutions:

Rosin solution
1 l  Water
10 g KOH
55 g Rosin

Alum solution
1 l Water
50 g Alum

Shake 100 cc Rosin solution with 55 cc Alum solution in a bottle.
A white precipitate will form. Pour it onto filter paper in a funnel.
When most of the water has drained, add more water to wash the resin.
Repeat this 3-4 times.
Spread the resin on a glass plate to dry.
This resin should now dissolve in Spirits Of Turpentine:

1 g resin
3 cc Spirits Of Turpentine

Stir this mixture continually until it's dissolved.
Then add to it

6 cc base varnish

Stir again and filter. You will need a very "open" filter, for example a hankerchief.
This clear varnish should be used at once, before it gels.
It spreads like a dream....!
But it needs strong sunlight to dry. Here in Norway a UV-light box is needed.

Making resin with color:

Extracting color from Madder:
To get most of the color out of the Madder, it should be fermented.
Powdered Madder is put in 1% Acetic Acid solution and permitted to stand for a couple of days, then left to dry thoroughly.
100 grams of fermented Madder in 500 cc of Alcohol can now be brought to the cooking point in a water bath, and then left to cool.
After filtering off the madder, put the colored alcohol back in the water bath to cook off largest part of the alcohol.
The colored Alcohol can now be shaken together with the Rosin and Alum solution to make colored resin.

Use the colored resin just like the uncolored resin.

Experiment to get the right color..
It will be orange or reddish on the finished instrument. If a few percent of the Alum is substituted with Ferrous Sulfate, it will be more brown.

This explanation can indeed be more detailed. If/when you run into trouble:
Please give comments or ask questions, and I will add details where needed.