NET DETERIORATION IN LAKE ERIE 169 



very seriously before. The mathematical treatment apparently is not 

 valid. One of Terada's first assumptions is that there is a constant 

 bacterial environment to serve as a source of weakening influence. 

 This is not true for any great length of time and occurs under condi- 

 tions where the bacterial population probably will lose its virulence 

 (juickly. Tauti assumes in his paper that bacteria increase at a rate 

 equal to a 77 +^ where a and ^ arc constants. This assumption is 

 true only when the bacteria are in the "logarithmic growth phase," 

 a condition which probably does not prevail in nature and only under 

 special conditions in the laboratory; hence further deductions can 

 not be held valid. 



Finely divided solids, such as the pigments used in the net preserv- 

 ative treatments, or mud will keep fibers from sliding upon themselves 

 just as wetting will and hence cause an apparent increase in the 

 strength of test twines at early times in the test period. As a result 

 many preservative treatments appear to increase the strength of the 

 twine during the time of the test — an increase in strength that does 

 not necessarily parallel the strength of the individual fiber — thus 

 making the experimental findings difficult to interpret. For this 

 reason it seems best that observation at this time be limited to exami- 

 nation of the amount of fouling, or at most a microscopic examination 

 of the individual fiber, and that the tensile strength determinations be 

 delayed until deterioration is well advanced. At that time many 

 samples should be taken, and the strengths of the twine carefully 

 determined. Only at this time, when the effect due to rotting is greater 

 than the extraneous effects due to felting, or the presence of solid 

 matter, can one get a good measure of the breakdown of the individual 

 fiber. Experience up to the present time seems to show that early 

 evaluation of net preservative treatments is apt to be misleading. 



The upper group of curves in Figure 10 represents mixtures using 

 cuprous oxide as a toxic material, and the lower group represents 

 those treatments using insoluble copper salts of organic acids for 

 this purpose. It will be noted in both cases the copper content de- 

 creases with time and approaches a fairly constant value. This is 

 most marked in the case of treatments 2815 and 2818. The curve for 

 treatment 282 is shown on this graph on a reduced scale in order to 

 economize space. 



An examination of these curves indicates that in most cases the 

 analytical determinations agree with each other best after the expira- 

 tion of the first period, in which the rate of decrease in copper concen- 

 tration is so much greater than it is during the second period, which 

 arbitrarily may be said to date from about the fourth week. The 

 only exception to this, apparently, is the case of treatment 282, which 

 in this series, as well as in others, has shown itself to be most erratic in 

 behavior. This treatment uses petroleum asphalt as a vehicle. This 

 material, therefore, should be dissolved in the water-gas-tar oil and 

 stirred until solution has progressed as far as possible. If a sluge 

 remains the solution should be poured off before the cuprous oxide is 

 added in order to minimize the segregation. Petroleum asphalt need 

 not be used except where coal tar is not easily purchased, but will 

 give results substantially as good as the other tars. Treatment 2825 

 likewise shows erratic results during the early part of the test. This 

 can be explained on the basis that the treatment was applied in a 

 "spotty " manner. This fact was quite obvious at the time the sample 



