during the previous week, approximately October 14.  These were left the longest on the bush of any hazel and still were not ripe although they were not entirely killed by the several frosts occurring before that time.  They are always much later than the wild hazel.”

On October 20, I had an opportunity of comparing the action of frost on the leaves of these plants.  Those of the White Aveline type had not changed color and were very green.  The leaves of the Jones hybrid showed some coloration but nothing to compare with those of the Winkler hazel, many of which had the most beautiful colors of any of the trees on the farm—­red, orange and yellow bronze.  Hazilbert No. 1, which resembles a wild hazel in appearance and habits of growth, had colored much earlier in reaction to the frost and was as brightly tinted as the wild hazel and Winkler plants except that, like the wild hazel, it had already lost much of its foliage.  Some of the wild hazels were entirely devoid of leaves at this time.  Hazilbert No. 5 showed the best color effects with No. 4 second and No. 2 last.

The color of the leaves and the action of the frost on the plants during the autumn is another thing, in my opinion, that helps to differentiate between and to classify European filberts, American hazels and their hybrids.  My conclusion in regard to the effect of frost is that the reaction of the Winkler hazel is very similar to that of the wild hazel in color but exceeding it in beauty since its leaves do not drop as soon after coloring.  At this time, the leaves had not changed color on the imported European plants, the Gellatly filberts from British Columbia or the White Aveline type.  They had turned only slightly on the Jones hybrid.  I think an accurate idea of the general hardiness of a plant is indicated by the effect of frost and by early dropping of leaves, using the sturdy wild hazel as the limit of hardiness and assuming that its hardiness is shown by both degree of coloration and early dropping of leaves.

In noting the action of frost on the Winkler hazel, I have mentioned that it was more like that on the American hazel than on the European filberts.  The Winkler has always been considered a native woodland hazel, but, although it does show several similarities to Corylus Americana, I have also noticed certain qualities which definitely suggest some filbert heritage.  I have based my theory on a study of the Winkler hazels which have been bearing annually at my farm for six years, bearing more regularly, in fact, than even the wild hazels growing nearby.  My comparisons have been made with wild hazels in both Minnesota and Wisconsin and with European filberts.