572 SUMMARY OF CURRENT RESEARCHES RELATING TO 



boundary of the yolk. As development goes on, this is replaced by a 

 secondary endoderm of mesodermic origin. Anteriorly this arises from 

 lateral sub-cesophageal bodies ; in the rest of the body from mesodermic 

 aggregation in the middle line (the "chorda" of Nusbaum, the "blood- 

 cell-strand " of Heymons). 



Glands of Caterpillars.* — L. Bordas has investigated a large series 

 of caterpillars in respect to their glands. He describes a series of silk- 

 glands of various types, and the accessory glands of Lyonnet. The 

 minute structure is also discussed, and the production of silk. Bordas 

 confirms what Gilson noted, that the raw material of the silk is elaborated 

 both in nucleus and cytoplasm. It passes into the internal cavity of the 

 gland, going through the cuticular intima by a process more like filtra- 

 tion than osmosis. In the lumen of the gland it undergoes various 

 changes. The thread of silk has a hyaline homogeneous core, the silk 

 proper, and a delicate cortical sheath, which is removed by washing. 

 The accessory glands secrete a liquid, or slightly viscous material, which 

 serves to unite the threads of silk, and possibly acts on them chemically, 

 so that they are rapidly hardened. Bordas has found thoracic glands 

 on most of the caterpillars he has studied. They sometimes secrete an 

 offensive fluid, as in the caterpillar of the puss moth. In almost all 

 cases there are mandibular glands, whose secretion is partly digestive, 

 but especially defensive. 



Thysanura and Collembola of Midland Plateau. f — Walter E. 

 Collinge enumerates two species of Thysanura and thirty-eight species 

 and three varieties of Collembola. One of the varieties of Collembola 

 is new, Lepidocyrtus lanuginosus var. plumbeus. 



Thorax of Insects and Articulation of Wings. :j: — R. E. Snodgrass 

 seeks to show the unity of structure that prevails throughout all the 

 orders of Insects. He accepts six primitive head-segments, provided 

 that the apparent superlingual is a real segment, and one microthoracic, 

 or neck-segment. The sclerites of the microthorax form the cervical 

 sclerites of the neck, often reduced or rudimentary, and the gular-plate 

 of the head, when such a plate is present. Its appendages form the 

 labium. 



The thorax proper has three segments. (In Hymenoptera the tergum 

 of the first abdominal segment is added.) These three segments are 

 primary metameres. The sclerites in each are subdivisions of the wall 

 of one primitive segment, and the apparent double nature of each seg- 

 ment is secondary. Characters that have been urged as special evidence 

 to the contrary, such as the equivalence of the episternum and epimerum 

 and the double structure of the coxa in some orders, lose their signifi- 

 cance when nymphal, larval, and pupal forms are examined. The wing- 

 is hinged to the notum on the two notal wing processes, and is sup- 

 ported from below upon the wing process of the pleurum. 



* Ann. Sci. Nat., x. (1909) pp. 125-98 (3 pis. and 22 figs.). 



t Birmingham Nat. Hist. Phil. Soc.,1910, 14 pp. (6 figs.) 



X Proc. U.S. Nat. Museum, xxxvi. (1909) pp. 511-95 (30 pis.). 



