SELECTING GIRLS AS ASSEMBLERS 55 that if they indicated the necessary qualities in men they would indicate them in women as well. However, this was impossible because the process of assembling had it self been entirely rearranged to meet the conditions of greatly increased quantity production. Hitherto, guns had been assembled in their entirety by one man. Now >t was proposed to break up this operation into a large number of steps, so that one assembler assembled one small part, another some other part, and so on, up to the 'ast step which consisted of putting together into the fin ished product all the parts that had been assembled. The °ld operation required a mechanic of considerable skill a nd long training. It will readily be seen that the new method did not require nearly the same degree of mechani cal skill, although it did require more speed and nimble- n ess in assembling a large number of relatively small and simple parts. Because of this change in the method of assembling, a standard on which to base the significance °f the tests was lacking. And consequently, the finding °f significant tests had to be something of a leap in the dark. The first step was to go into the shop and examine the ex act nature of the proposed operations. It was not con- sidered sufficient to do this by asking questions and ob serving the work from a little distance. A first hand knowledge was obviously the best, and therefore the ex perimenter actually performed the various operations at a bench and in the customary way, until he was able to do them with some degree of celerity. Instead of de- scribing this work in technical minuteness, it will be best km the purposes of this exposition to describe it in terms °f the tests which were finally devised. The work of assembling gun parts as it is now done