No. 504] ORIGIN OF FROTH IN SPITTLE-INSECTS 795 



water lived for some time and he concludes that the above- 

 referred to structures have a respiratory function. These 

 so-called branchial tufts of Morse (sc., Figs. 1 and 2) are 

 nothing more nor less than the plates of secretion from 

 the glands described above. This secretion is not fila- 

 mentous, but appears rather very granular. It is very 

 easy to remove these flakes of wax by means of camel's 

 hair brush from the segments to which they adhere but 

 slightly. The secretion is not soluble in alcohol and 

 swells up after being allowed to stand in water for some 

 time. It has a very striking superficial resemblance to 

 beeswax. 



The resemblance of these glands to the wax-glands of 

 the honey bee, in the shape of the cells and in the location 

 and appearance of the nuclei is very striking. This fact 

 has already been pointed out by Batelli (1891), Gruner 

 (1901) and others, who also find much in common with 

 the cells of the wax-glands of coccids and other Hemip- 

 tera. Gruner (1901) thought that the wax served to 

 line the "pocket" already mentioned so as to enable the 

 air to penetrate it the better, and prevent the inflow of 

 the fluid while the bubbles of air were being blown into 

 the fluid. Porta (1900) accepting Morse's curious error 

 regarding the presence of branchife, considered this thick- 

 ened epithelium as a supporting structure for these sup- 

 posed gills. Carrying further Batelli 's misinterpreta- 

 tion of Wheeler's work, he regarded these "supporting 

 structures" as homologous with the abdominal append- 

 ages of the embryos of Nepa and Cicada. 



I am convinced that these glands secrete a mucil- 

 aginous substance which is utilized by the insect in the 

 production of the spittle. Mixed thoroughly with the 

 anal secretions it serves to render this viscid and thus 

 to hold the air bubbles blown into it. This view is based 

 upon the following experiments. 



The region of the seventh and eighth abdominal seg- 

 ments of several specimens of the nymph was carefully 



