Another important point in hickory propagation work consists in the employment of the Spanish windlass for fastening graft and stock together. The old time wrapping of twine or of raffia had to be released in order to allow growth at the point of union of scion and stock. When cord is used it cuts deeply into the new growth, and raffia, which is placed on flat, will be burst open. In either case new wrapping is required at a precarious time, according to old methods. The Spanish windlass, which is used in surgery for controlling haemorrage, seemed to me to be applicable for fastening scions in place. It consists in a paraffined cord with ends tied in a firm knot but hanging loosely about the graft and wound. A wooden skewer or any small lever, is then inserted into the loose loop of cord and twisted about until the part of the cord about the graft wound is so snug that it holds the scion in place more firmly than it can be held by any other sort of wrapping. In order to prevent the cord from cutting into the bark, two shields of wood or metal an inch in length, are interposed between cord and bark. The lever of the Spanish windlass is fastened with a cord or with a galvanized nail in order to prevent the windlass from unwinding and the whole covered with melted paraffin. This may remain in place for two seasons without change, holding the scion firmly in place all of that time and requiring no attention. The growing stock separates the two shields very much as it might separate two stones in the field and automatically unwinds the Spanish windlass by sheer force, just enough to allow growth without any unloosening of its holding apparatus. In hickory grafting, much experimental work remains to be done in the choice of stocks for grafts of different species. Almost all of the hickories that have been grafted upon the pecan hickory stock, seem to do pretty well upon that stock, but the converse is not true. The pecan apparently does not do well as a rule when grafted upon other hickory stocks, even upon those of its cousins in the open-bud group. The shagbark hickory, in my experience, has done best upon stocks of the shagbark or mockernut or pignut. A number of years, however, are required in some cases for determining that point. Shagbarks which I have grafted upon bitternuts have sometimes made a remarkably good start. Then at the end of three or four years they begin to slow up, while shagbarks on shagbark stock, starting slowly at first, surpassed the ones on bitternut stock finally.