The earlier studies seemed to indicate that the irritation was purely mechanical in origin, the result of the minute barbed hairs working into the skin in large numbers. Subsequently, however, Dr. Tyzzer (1907) demonstrated beyond question that the trouble was due to a poison contained in the hairs. In the first place, it is only the peculiar short barbed hairs which will produce the dermatitis when rubbed on the skin, although most of the other hairs are sharply barbed. Moreover, it was found that in various ways the nettling properties could be destroyed without modifying the structure of the hairs. This was accomplished by baking for one hour at 110° C, by warming to 60° C in distilled water, or by soaking in one per cent. or in one-tenth per cent. of potassium hydrate or sodium hydrate. The most significant part of his work was the demonstration of the fact that if the nettling hairs are mingled with blood, they immediately produce a change in the red corpuscles. These at once become coarsely crenated, and the roleaux are broken up in the vicinity of the hair ([fig. 37b]). The corpuscles decrease in size, the coarse crenations are transformed into slender spines which rapidly disappear, leaving the corpuscles in the form of spheres, the light refraction of which contrasts them sharply with the normal corpuscles. The reaction always begins at the basal sharp point of the hair. It could not be produced by purely mechanical means, such as the mingling of minute particles of glass wool, the barbed hairs of a tussock moth, or the other coarser hairs of the brown-tail, with the blood.
The question of the source of the poison has been studied in our laboratory by Miss Cornelia Kephart. She first confirmed Dr. Tyzzer's general results and then studied carefully fixed specimens of the larvæ to determine the distribution of the hairs and their relation to the underlying tissues.
The poison hairs ([fig. 37]), are found on the subdorsal and lateral tubercles ([fig. 38]), in bunches of from three to twelve on the minute papillæ with which the tubercles are thickly covered. The underlying hypodermis is very greatly thickened, the cells being three or four times the length of the ordinary hypodermal cells and being closely crowded together. Instead of a pore canal through the cuticula for each individual hair, there is a single pore for each papillæ on a tubercle, all the hairs of the papilla being connected with the underlying cells through the same pore canal, (figs. [39] and [40]).
