This paper describes some corona tests which were made in Peru on a 70-mile transmission line located at an average altitude of 13,300 feet. The voltages range from 40,000 to about 72,000. The method employed in the tests is described and the calculations are given. The corona losses were found by subtracting the core and copper losses of the transformers and the line copper losses due to the line charging currents from the total loss measured. Core-loss tests were made on the transformers in order to check these against the shop tests. The highest voltage experimented with was 72,300 when a corona loss of 153 kw. was found, or a little over two kw. per mile of line. The results are compared to ``Peek's law'' and a wide divergence found, amounting to over 800 per cent at about 66,000 volts and a loss of 58 kw. Close correspondence is found between these tests and those made by Faccioli on the Shoshone-Leadville line, and the equation of the loss-voltage curve obtained is the same as was established by Faccioli, and shows this relation to be logarithmic and not quadratic for the voltage range investigated. It is shown that the results are in good agreement with the formulas deduced by Professor Ryan from extensive laboratory tests. Finally test data are given for a test made during rainy weather and this shows no appreciable difference from tests made during fair weather.