176 A. J. Eivart: 



A, contains very nearly the theoretical amount of Ca formate 

 required for the production of pentose sugar (1 mol. Ca formate = 9 

 molecules ^'H^O). If any hexose or disacc.haride is formed, less 

 calcium formate would he required for optimal polymerisation. 

 The higher polymerization in B indicates that some. hexose or di- 

 saccharide is produced. With the intermediate concentrations, the 

 results are very consistent, varying less than 1 % in duplicate tests. 

 At the extremes, however, the results obtained particularly at the 

 lower extreme are apt to vary somewhat, liowever, carefully the 

 tests are performed. 



Evidently, too great an excess of calcium formate interferes 

 slightly with polymerisation at the higher concentrations, whereas 

 in lower concentrations the opposite effect appears to be exercised. 

 The maximum polymerisation is given with concentrations corre- 

 sponding to 0.62% to 0.75% solutions of formaldehyde. 



When strong solutions of alkali are used , or when the formalde- 

 hyde is concentrated, portion of the alkali is apt to attack the 

 sugar produced, decreasing the apparent polymerisation. 



When calcium formate is present, however, and the soda is added 

 gradually, the percentage polymerisation decreases less rapidly 

 with increasing concentration. 



Thus, adding 10% c.c. of 31.5% CH^O to 50 c.c. of 10% CaCa^ 

 (6.3% CH.O), 28.8 c.c. of 3.5% Na HO was required to complete 

 the reaction, and an abundance of sugari was formed. In this case 

 52% of the 6% CHvrO was polymerised. Using a mixture of 50 

 c.c. of 10% CaC0^2^ ai^^ 20 c.c. of 31.5% CHgO, and running in 

 35% NaHO, it is necessary to obtain an approximately accurate 

 result to keep the temperature below boiling point, and to shake 

 vigorously after each addition of soda until the precipitated lime 

 dissolves as formate. Even then the liquid becomes distinctly 

 brown before all the formaldehyde has been removed. The amount 

 of soda averaged 6.6 c.c, and 1 c.c. of CHgO equalling 0.6 c.c. of 

 soda, the percentage polymerisation was 45. 



Polymerisation by other divalent alkaline metals. 



In the previous paper it was found that the presence of neutral 

 barium, strojitium or magnesium salts in boiling formaldehyde, to 

 which caustic soda was added, increased the amount of polymerisa- 

 tion, and decreased the amount of soda required to neutralise the 

 formaldehyde. Barium and strontium were not quite so effective 

 as calcium, and magnesium had comparatively small effect. In 

 these preliminary tests the importance of having a slight excess of 



