xxxiii, '221 ENTOMOLOGICAL NEWS 39 



and may have to do with forcing the hlood into the penis when the 

 latter is erected. 



Fig. 24 is section G through the proximal vesicle, pi', and the proximal 

 meatus, />;. It shows also the wrinkled and unexpanded condition of 

 the penis shaft. 



Fig. 25 is section H through the proximal meatus and below the proxi- 

 mal vesicle. It also passes through the attachment of the penis to the 

 sternum and shows the tissue of the body-cavity continuous with the 

 penis cavity. The author has found no actual opening from the ab- 

 domen into the base of the penis hut he has had only dried and poor 

 material to work with. It is of course remotely possible that the 

 Libellula penis is never truly erected, that the erections produced by 

 boiling the specimens, as in fig. 26, may be wholly unnatural. 



In the series of species beginning with the primitive com- 

 posita and terminating in such specialized forms as librans 

 and iucesta, the softer parts of the penis are covered with a 

 dense plush of hairs, which become erect when the penis is 

 distended. These can usually be demonstrated by boiling the 

 penis, if the material is not too old. Fig. 27 shows the penis of 

 jcsscana relaxed and fig. 26 the same organ after boiling, when 

 it is supposedly erect. Fig. 28 is a cross-section through the 

 penis of incest a at the line shown in fig. 10 and marked sec. 

 Fig. 29 is the upper part of fig. 28 enlarged. The erectile hairs 

 shown in these figures are hollow outgrowths of the soft cuti- 

 cula and fill and become erect when the main organ is distended. 

 At that time they are distended and their membraneous base 

 is evaginated, so that the boiled incest a penis appears somewhat 

 as does the erect jcsscana penis. When the penis is relaxed 

 these hairs are so completely withdrawn that their presence may 

 be entirely unsuspected. 



This paper has shown how little is known definitely concern- 

 ing the genitalia in the Odonata. The next paper will show 

 how useful the genitalia are in indicating the relationships 

 within the genus Libellula. 



BIBLIOGRAPHY. 



RATHKE, 1832. De libellarum partibus genitalibus. 



Gom>AKi>. IS'id. On the second abdominal segment in a few /.//'. ////</</:. 



Proc. Amer. 1'hil. Soc. XXXV, i>i>. -'05-212. 

 THOMPSON, 1 ( >OS. Appendages of the second abdominal segment of male 



dragonflics. Hull. X. Y. State Mus. No. 124, pp. 249-263. 

 BACKIIOFF, 1910. 1 >ie Kntwicklung des Copulationsapparates \<>n Agritm. 



Xiitschr. wiss. Zool, XCV, pp. 647-/Ut.. 



