168 U. S. BUREAtJ OF FISHERIES 



By using the modified Kjeldahl digestion method the analyst is 

 able to handle many samples, the number depending on the facilities 

 available for Kjeldahl digestions. The process of digesting requires 

 the minimum of attention and the analysis of the prepared solutions 

 can be conducted during its course. 



Some samples, notably those which contain the most tar, took as 

 long as two hours to digest ; but most of the samples required about 

 one-half hour. By planning the work carefully it was possible for 

 an analyst using this procedure to analyze 10 to 15 duplicate samples 

 a day, making a total, of 20 to 30 determinations. This progress 

 would be impossible if each sample had to be destroyed by ashing. 



The tabulated results are shown in Table 4. The results of the 

 analytical determinations are plotted in Figure 10 with the copper 

 content of the tmne in milligrams per lineal foot as abscissae and the 

 time in weeks as ordinates. The data for the tensUe-strength deter- 

 minations are not plotted, for the reason that the period of the test 

 was so short that no clear-cut differences are apparent early in the 

 test and since the change of tensile strength with time gives no very 

 valuable information. 



The treatments tested in the lake were as follows: 



281. Untreated twine. 



282. 6 per cent cuprous oxide, 20 per cent petroleum asphalt, 74 per cent water 

 gas tar oil. 



284. 6 per cent cuprous oxide, 20 per cent coal tar, 74 per cent gas tar oil. 

 287. 6 per cent cuprous oxide, 20 per cent coal tar, Mo per cent mercuric oxide, 

 74 per cent water gas tar oil. 



2811. Two-thirds coal tar, one-third wood tar, apphed warm. 

 2815. 29 per cent copper oleate, 71 per cent kerosene. 



2817. 29 per cent copper resinate, 71 per cent water gas tar oil. 



2818. 15 per cent copper oleate, 50 per cent coal tar, 35 per cent benzol. 



2819. y^ per cent mercuric oxide, 6 per cent ferric oxide, 13 per cent coal tar, 

 7 per cent wood tar, 73 per cent water gas tar oil. 



2825. A commercial preparation, said to contain copper salts of naphthenic 

 acids. - — - 



The apparent increase in strength of twine when tested immediately 

 after immersion in water is disconcerting at first sight. It has been 

 noticed before and ascribed to errors of sampling, or variations in 

 the strength of the twine whereby occasional lengths of twine stronger 

 than usual would be encountered. Such is not always the case. 

 Larger twine especially shows a tendency to "felt" or more properly 

 to have its fibers become entangled more than usual after immersion 

 in water, with the result that individual fibers do not slip as easily 

 upon each other and that more fibers are broken than is ordinarily 

 the case when the whole twine breaks. The apparent strength is, 

 therefore, higher than before even though the fibers making up the 

 whole may individually be weaker at the time of the test. 



A number of Japanese investigators have studied this phenomenon. 

 Terada (1924) has shown that the weakening of twine immersed in 

 water is dependent upon temperatures and is greater in warm water 

 than in cold. This investigator also studied the rate at which twine 

 decomposed and expressed the results in a complicated equation. 

 Tauti (1927) has studied this question further and presents additional 

 experimental data and mathematical analyses. This investigator's 

 paper, and the preceding one, seem to be of most value in that they 

 call attention to the fact that the process of rotting is a complicated 

 one, and that there are facts involved that have not been considered 



