426 Dr. Heineken on the lieproducfum of 



count thirty, the number of the other is indistinct. — Killed 

 for a preparation. 



2. 

 July 25. Two pupse of Blatta Maderce. — Both antennae removed near 



the base. 

 Aug. 27. Have moulted and reproduced similar antennae to those of 1 , 



— One kept alive, the other killed for a preparation. 

 Sept. 27. Antennae as before. I can count 28 joints in the one, and 17 



in the other : the basal and third joints of the two do not 



correspond in length. 

 Oct. 25. Antennae nearly in the same state. It has not again moulted. 



3. 

 Aug. 8. A pupa of Reduvius ? Both antennee divided through 



the basal joint. 

 Sept. 4. Reproduced, at the same time moulting into a perfect insect. 



They are of unequal lengths, thicker and shorter than the 



original ones, and, as far as T can ascertain, having only 



three instead of four joints. 



27. The same. 



Oct. 10. Died from want of proper food, the antennae still unequal 



and imperfect. It had not moulted again. 



In the above instances sufficient care was not taken to leave the stumps 

 of both antennae of equal lengths, or to note accurately the joint at which 

 the amputation took place, and the number of joints which each antenna 

 possessed previously, to explain why the reproductions were unequal and 

 imperfect — some future experiments, conducted with more precision, 

 may perhaps elucidate it. I had previously and long ago removed the an- 

 tennae, one or both, from many perfect Blattce, Forficulce, Grylli, Lo- 

 custa, Acrydia, &c., but although they survived for different periods, 

 from a fortnight to even two months or more, reproduction never took 

 place ; it is, therefore, I think, clear that they are amenable to the same 

 conditions as spiders. I intend to pursue the inquir as extensively as 

 my feeble means will allow, upon the larvae and pupae of insects gene- 



