They were dipped in a dish of melted parafin wax for an instant and then quickly shaken in the air before scorching could occur. The scions were then grafted into a small chinquapin stock. A few days later one of the larger leaves of the larger shoot had cleared itself from the wax coating and had begun to expand widely, turning to a natural green color. The stem of the shoot turned to a normal brownish red. Two tiny shoots then broke through the wax of the larger shoot, looking like axillary bud shoots until closer examination showed them to be scale bud shoots. That should interest plant physiologists. Eventually the cramped leaves remaining under wax coating that was unnecessarily dense finally dropped away useless. The single green leaf and the two scale bud shoots went on to natural development. The smaller shoot of the other scion managed to burst through the wax completely and made normal growth. After these scions were well under way I went out and searched in the loose dirt and leaves of the old heap and found another hybrid chestnut scion that presented the allusive emblem of a canary bird. This one had a shoot of about half of one inch in length and it burst completely through the wax, to make a fine little twig. So much for an experiment that led immediately to one of far greater importance. If canary bird shoots could be made to break rules of horticultural theory and of recorded fact perhaps we might note the principle and apply it to the experimental grafting of green shoots of the year in tree propagation. This is what lawyers might call a _non sequitur_. Such grafting had always been a failure so far as I knew, and certainly my own attempts had failed in former years. Grafting of new growth of the year upon new growth of the year in the growing season is an established feature of horticultural experiment with certain annual plants. Why had it so signally failed with perennial plants and most impressively with trees? Doubtless plants produce in their leaves a hormone which directs certain enzymes that conduct wound repair by cell division. If plants which do not lignify for winter manage to direct successful wound repair after grafting and if plants which do lignify for winter do not conduct successful repair of grafted new growth it occurred to me in a speculative way that the reason might perhaps be sought in the nature of the two different kinds of hormones or of enzymes belonging to annuals and to perennials respectively. The difference might possibly depend upon the arrangement of ions, anions and cations upon two sides of the permeable membrane of a repair cell. The cell is an electrolyte and therefore division of the cell in course of preparation for multiplication might perhaps depend upon an electric impulse so delicately in balance that Nature for some cryptic reason might prefer not to allow the necessary balance to go toward cell division in grafts consisting of green growth of the year in perennials.