281 [rackard. 



ing the ovipositor, lie in separate pairs, in two groups, exposed dis- 

 tinctly to view. The ovipositor thus consists of three pairs of slender 

 non-articulated tubercles arising on each side of the mesial line of the 

 body in juxtaposition. The first two pairs arise from the eighth 

 abdominal ring, and the third pair grow out from the anterior edge 

 of the ninth ring. The ends of the first pair scarcely reach beyond 

 the base of the third pair. With the growth of the semi-pupa, the 

 terminal or tenth ring decreases in size, the tip of the abdomen is 

 gi-adually incurved toward the base, (Fig. 2), and the three pairs of 

 rhabdites approach each other so closely that the two outer ones 

 completely ensheath the inner, until a complete distensible tube 

 is formed, which gradually is withdrawn entirely within the body (see 

 Fig. 4). The male genital organ is originally composed of three pairs 

 of non-articulated tubercles all arising from the ninth abdominal ring, 

 being sternal outgrowths, and placed on each side of the mesial line of 

 the body, two being anterior, and very unequal in size, and the third 

 pair nearer the base of the abdomen. Thus in their position, the 

 three pairs of tubercles destined to form the male intromittent organ 

 can not be said to be strictly homological with the female ovipositor; 

 nor can the external genital organs be considered as in any way homo- 

 logous with the limbs, which are articulated outgi'owths budding out 

 between the sternal and pleural pieces of the arthromere*. This 

 view will apply to the genital armor of all insects, so far as I have 

 been able to observe. It is so in the larva of Agrion, which com- 

 pletely repeats the structure of the ovipositor of Bombus in Its essen- 

 tial features detailed above. Thus in Agrion the ovipositor consists of 

 a pair of closely appressed ensiform processes which come out from 

 under the posterior edge of the eighth abdominal ring, and are em- 

 braced between two pairs of thin lamelllform pieces of similar form 

 and structure, arising from the sternlte of the ninth ring. These ster- 

 nal outgrowths do not homologize with the long filiform antennae-like, 

 jointed appendages of the tenth ring, as seen in the Perlldas and 

 most Neuroptera and Orthoptera, which, arising as they do from 

 the arthropleural, or limb-bearing region of the body, i. e., between 

 the sternum and episternum (or lower pleurite) are strictly homolo- 

 gous with the abdominal legs of the ]\Iyriapoda and the "false legs" of 

 caterpillars. So that in these genito-sensory appendages, we perceive 

 faint tracings of the idea of antero-posterior symmetry first observed 

 in vertebrates by Oken, and more recently by Professor Wyman, 



*This term is proposed as better defining the ideal ring, or primary zoological ele- 

 ment of an articulate animal than the terms somite or zoUnite, which seem too 

 vague; so also the termarfhrorlerm for the outer crust or body walls of articulates, 

 and arthropleura for the pleural or limb-bearing region of the body, being that por- 

 tion of the arthromere comprised between the tergite and sternite. 



