476 TEXT-BOOK OF ENTOMOLOGY 



that Machilis everted its sacs when the vessel in which it was put was filled with 

 warm, damp air. The sacs are only thrust out when the creature is completely 

 at rest. 



Structures referable to blood-gills also occur temporarily in the 

 embryo of Orthoptera; Rathke observed them in the mole-cricket; 

 Ayres observed them in CEcanthus niveus, where they form tw r o 

 stalked broad oval appendages on the first abdominal appendages, 

 which he regarded as gills. Patten observed them in Phyllodromia 

 germanica, as pear-shaped structures occurring in the same situation, 

 but regarded them as sense-organs, as did Cholodkovsky. Graber 

 found these structures in the embryo of the May -beetle, which looked 

 like the other embryonic limbs, but survived after the disappearance 

 of the latter, being longer and broader and unjointed. These disap- 

 peared shortly before birth. In Hydrophilus they remain, Graber 

 states, after birth. Nussbaum has seen them in Meloe. 



Finally, Wheeler has discussed at length these embryonic organs, 

 which he regards as glandular structures, and calls plenropodia, their 

 primitive function having been that of limbs. He has detected them in 

 the embryo of Periplaneta orientalis, Mantis Carolina, Xiphidium ensi- 

 ferum (Fig. 387) ; also in the Hemiptera (Cicada septemdecim, Zaitha 

 Jluminea), and in Sialis infumata. He discards the view that they 

 were once gills or sense-organs, and concludes that they were glands. 

 But, as we have suggested, their function once that of gills, and still 

 respiratory in Synaptera, has perhaps become in the winged insects 

 glandular and repugnatorial. Instead, then, of being modified 

 abdominal limbs afterwards serving as glands, as Wheeler claims, 

 we are inclined to believe that they functioned as blood-gills. 



A". Tracheal gills of adult insects 



Tracheal gills are known to be retained by a few insects in the 

 imago stage, the nymphs in all stages breathing by them. The most 

 notable example is the perlid genus Pteronarcys, in which, as New- 

 port states, there are eight sets, comprising 13 pairs of branchial 

 tufts distributed over the under surface of the thoracic and first two 

 abdominal segments. 



The first set, consisting of three pairs of tufts, partly encircling the neck like 

 a ruff, arises from the soft membrane connecting the head and prostermun. 

 The thoracic tufts originate between and behind the COXEE, as well as on the 

 front margin of the meso- and metathoracic segments. The number of filaments 

 in each tuft varies from about 20 to 50 or more, the densest tufts being those of 

 the two hinder thoracic segments. Each filament is usually simple, though in a 

 few cases they are branched (Fig. 457, A). 



