"Some time ago I worked with several colleagues in the laboratory of a high school on a series of complicated elasticity experiments, a piece of work which we had undertaken voluntarily but which began to take more time than we had expected. One day as I went into the laboratory with my colleague F., the latter remarked how unpleasant it was to him to lose so much time that day, since he had so much to do at home. I could not help agreeing with him, and remarked half jokingly, alluding to an incident of the previous week: 'Let's hope that the machine gives out again so that we can stop work and go home early.' "In the division of labor it happened that F. was given the regulation of the valve of the press, that is to say, he was, by means of a cautious opening of the valve, to let the liquid pressure from the accumulator flow slowly into the cylinder of the hydraulic press. The man who was directing the job stood by the manometer (pressure gauge) and when the right pressure had been reached called out in a loud voice: 'Stop.' At this command F. seized the valve and turned with all his might--to the left! (All valves, without exception, close to the right.) Thereby the whole pressure of the accumulator suddenly became effective in the press, a strain for which the connecting pipes are not designed, so that a connecting pipe immediately burst--quite a harmless defect, but one which nevertheless forced us to drop work for the day and go home. "It is characteristic, by the way, that some time afterward when we were discussing this occurrence, my friend F. had no recollection whatever of my remark, which I could recall with certainty." From this point you may reach the conjecture that it is not harmless accident which makes the hands of your domestics such dangerous enemies to your household property. But you can also raise the question whether it is always an accident when one damages himself and exposes his own person to danger. There are interests the value of which you will presently be able to test by means of the analysis of observations.