NOTES AND MEMORANDA. 57 



even less truly conical ; eacli consists of a conical or rather globular 

 head, a long central filamentous portion, and a proximal spiudle-shaj)ed 

 portion. There is no cornea in either the lateral or the frontal eye, 

 but the distal end of each cone comes into relation at the surface with 

 a double cell, containing two (Semper's) nuclei. These double cells 

 do not fit closely to one another, but leave triangular interspaces : the 

 boundary wall between the two halves of which they are composed 

 penetrates into a very evident cleft, marking the division of the cone 

 into two longitudinal segments. Schmidt's observations give no 

 support to Pagenstecher's view that this sej^aration of the cone into 

 two longitudinal moieties is an evidence of multiplication by division. 



But perhaps the most important observation on these cones is 

 that in hardly a single case is the axis of the visual rod at right 

 augles to the external or corneal surface, so that Miiller's theory of 

 mosaic vision is here quite inapplicable, since there is neither the 

 straightness of the refracting bodies, nor the contrivance for absorp- 

 tion of lateral rays required by that view of the action of the so- 

 called compound eye. The author considers that the eyes o{ Phronima 

 are mere makeshifts for image-forming organs, and that they servo 

 only to distinguish different degrees of light and colour. 



Observations on the visual rods of other Crustacea showed that in 

 Palcemon many of the cones are straight, but that those at the periphery 

 of the organ are oblique to the corneal facets, their proximal segments 

 being strongly bent. In Palinurus this flexure sometimes amounts to 

 90°. In the lobster the rods are very irregular, hardly two being 

 alike : their j^roximal segments show the greatest amount of vari- 

 ability as to size and degree of flexures, and have no resemblance at all 

 to image-forming bodies. 



The only insect examined by the author is Dytiscus marginalis ; in 

 it, as in the prawn, he finds that the rods towards the perijihery of the 

 eyes exhibit a marked flexure. The paper is accomjianied by a plate. 



Poison Glands of the Centipedes. — It has long been known that 

 the Chilopod Myriapoda, commonly known as centipedes, which are 

 carnivorous in their habits, kill their prey by a poison injected at the 

 first bite of their formidable nippers. The seat of the glands secreting 

 the poisonous fluid was, however, unknown, the organs formerly suj)- 

 posed to secrete the venom being found to j)oui- their secretion into 

 the cavity of the mouth and not into the nijjpers. Mr. McLeod, 

 during a residence in Java, examined some of the large centipedes 

 with which that island abounds, and esiJecially Scolopenclra horrida, 

 and finding the glands which might easily be taken for poison glands 

 had nothing to do with the nippers, which nevertheless always ex- 

 hibited a very distinct oritice at the tip, he was led to search for the 

 glands in the interior of those organs themselves. 



The process he adopted is one that has of late given admirable 

 results in the investigation of the anatomy of many animals ; namely, 

 the preparation of sections of them in various directions, after they 

 had been immersed in melted paraffin, the subsequent hardening of 

 which keeps all parts in their natural positions during the ojieration 

 of cutting. By this means he detected the poison gland, which is 



