is reached the test-tube is brought to the edge of the flame, and the boiling is continued at the middle height of the flame (see fig. 8). In this way it is possible to carry out the boiling continuously and energetically, so that the liquid travels over more than half of the test-tube, without any danger of over-boiling. Not for a single moment should the attention be allowed to wander from this process, for
Fig. 8.
(a) Test-tube holder; (b) boiling-stick.
everything depends upon the accuracy of the operation. If the ebullition is too weak, then under certain circumstances the reaction may fail altogether, while, if the rate of ebullition differs in the different tests, we get a difference in the intensities of the colorations. The results, in short, are inaccurate.
After the lapse of half an hour a comparison of the
