By Madelyn Webb, H. Ruber, G. Leduc
Water Research, Volume 10, Issue 4, 1976, Pages 303-306, ISSN 0043-1354, DOI: 10.1016/0043-1354(76)90171-8.

Preliminary testing of eight collectors (xanthates) and four frothers in 96-h static and 28-day flow-through bioassays using rainbow trout as the test organism show a great disparity in the toxicity of the chemicals administered in these two ways.
For the short-term tests, the relative toxicity of the compounds is expressed as an lc50 or as a range of concentration in mg l-1 in which the lc50 is expected to fall. Of the collectors tested in this way sodium ethyl and potassium amyl xanthate were the most toxic, with lc50's in the range of 30-50 mg l-1. Among the frothers, xylenol (cresylic acid) was found to be the most toxic (5.6 mg l-1 >lc50 > 3.2 mg l-1) while polypropylene glycol was least toxic (lc50 > 1000 mg l-1).
The long-term tests using potassium ethyl, sodium isopropyl, sodium ethyl, and potassium amyl xanthate indicated that in the flow-through system, the toxicity of the chemicals was in the order of 100 fold greater compared with the static bioassay results.

More details...