The thermometers that Welsh used were undoubtedIy far superior to any that were devised before his time and it is much to be regretted that they were allowed to fall into disuse. Perhaps the most important stricture on the observations that will have to be recorded is that the observers were not provided with a base station, on which account the value of results was impaired. It was not realised that it was necessary to make observations on the ground to compare with those that were being made at high altitudes.

Welsh made, in all, four ascents in the summer and autumn of 1852 and in his report he is careful to give the highest praise to his colleague, Green, whose control over his balloon he describes as "so complete that none who accompanied him can be otherwise than relieved from all apprehension, and free to devote attention calmly to the work before him."

The first ascent was made at 3.49 p.m. on August the 17th, under a south wind and with clouds covering some three-quarters of the sky. Welsh's first remark significant, and will be appreciated by anyone who has attempted observational work in a balloon. He states naively that "a short time was lost at first in an attempt to put the instruments into more convenient order, and also from the novelty of the situation." Then he mentions an observation which, in the experience of the writer, is a common one. The lowest clouds, which were about 2,500 feet high and not near the balloon, were passed without being noticed; other clouds were passed at different heights; and, finally, a few star-shaped crystals of snow; but the sun shone almost constantly. Little variation occurred in the direction of travel, which averaged thirty-eight miles an hour, and the descent took place at 5.20 p.m. at Swavesey, near Cambridge.