XI. A. 4 Witt: Testing of Galvanized Iron 151 



then in alcohol, after which it is dried and weighed. Or the 

 lead may be determined by dissolving in nitric acid and titrating 

 with potassium dichromate. 



COMMENTS ON THE LEAD ACETATE METHOD 



In the original paper it is recommended that a square (5.08 

 by 5.08 centimeters), which should not vary more than 0.04 cen- 

 timeter (%4 of an inch) plus or minus, be used for the test. 

 This is too small a piece to give accurate results under ordinary 

 conditions. If two such squares are cut so that on each side of 

 the first there is an error of plus 0.04 centimeter, and on each 

 side of the second an error of minus 0.04 centimeter, the total 

 error will amount to over 6 per cent. It is practically impossible 

 to cut sheeting as accurately as this with ordinary shears, and 

 a great many pieces are tested with errors much greater. Some 

 of the pieces cut by power shears are not sufficiently accurate." 

 Another and more serious reason against the use of such small 

 pieces is that they do not give a fair average of the composition 

 of the material. Any variation in the coating is, of course, 

 multiplied by thirty-six in calculating the results to the basis of 

 900 square centimeters. 



That the variation in coating may be considerable was dem- 

 onstrated by some tests made in this laboratory. The results 

 are show in fig. 1. The maximum variation was 2.19 grams 

 per 15-centimeter square, or about 12.3 per cent. This is equiv- 

 alent to 8.76 grams per 900 square centimeters, or about 0.3 

 ounce per square foot. It is easy to see that had only the small 

 squares been used a great error would have been possible, leaving 

 out of consideration any inaccuracies in cutting. Nothing 

 smaller than 15-centimeter squares should be used for testing 

 galvanized sheeting, and when convenient 30-centimeter squares 

 are to be recommended. All tests made in this work by the lead 

 acetate methods have been with 15-centimeter squares, so that 

 average results might be obtained. Such a procedure is satis- 

 factory for experimental purposes, but would be too expensive 

 for a laboratory where a great many tests are made. 



Theoretically, to remove the zinc from a 15-centimeter square of 

 galvanized iron containing 68.8 grams of zinc per 900 square cen- 

 timeters, about 100 grams of lead acetate [Pb (0,11302)2 • 3H2O] 

 would be required. In practice, however, it has been found 



" These considerations are intended to apply to commercial testing only. 

 Of course, for experimental purposes, it is possible to prepare squares 

 having negligible variations in size, but this requires considerable time and 

 is not practicable for routine work. 



142765 2 



