126 MEMOIES OF THE NATIONAL ACADEMY OF SCIENCES. 



the blind forms have the longer olfactory rods, and that they are developed so as to compensate 

 for the loss of vision. 



Si 110ns coniparons entre elles les baguettes de nos Amphipodes d'eau douce, c'est-a-dire celles des Gammarus 

 pulex, fluviatilis et puteanus et celles de V Asellus aquaticus et de la forme provenant des puits, nons verrons que ces 

 baguettes n'atteignent pas chez toutes ces especes le rueme d6veloppement ; celles des Gammarus pulex et fluviatUis 

 et celles de Y Asellus aquaticus sont courtes, atteignant a peine le tiers de la longueur de l'article suivant, tandis que 

 les baguettes du Gammarus puteanus et de P Asellus des puits egalent en longueur et merne d<5passent l'article suivant. 

 (Je me sers ici de l'article suivant comme mesnre comparative, vu que tous les articles des tigelles de ces especes sont a 

 peu pies e"gaux entre eux et que les baguettes, fix6es a 1' angle des articles, sont dirige"es parall element a l'article qui suit. ) 



Nous remarquerons ensuite que les baguettes courtes appartiennent aus especes pourvues d'organes visuels et 

 que les baguettes longues, que nous considererons comme tres d6veloppe"es, appartiennent aux deux especes habitant 

 leseaux souterraines et qui sont aveugles. 



Comparant d'une maniere g6n6rale l'importance du sens visuel et olfactif dans Pe'conomie animale, nous 

 verrons que l'un et l'autre sont d'une n6eessit6 e"gale, ou que si Pun est atrophi6, Pautre par contre s'est d^veloppe', 

 s'est perfectionne" de maniere a pouvoir remplacer Porgane deTectueux (I. c, p. 12). 



Afterwards Leydig, and also Fries, confirmed this view, and the former remarks in his Unter- 

 euchungen, etc., 1883: 



In two Crustacea of our native dark fauna it has been for some time possible to indicate such a change in rela- 

 tions." The olfactory rods on the larger flagellum of the upper antenna? in the eyeless Gammarus puteanus are more 

 developed than in species provided with eyes. Short, and scarcely reaching in Gammarus pulex and Gammarus fluvia- 

 tUis a third of the length of one flabellar-joint, they are in Gammarus puteanus nearly of the length of the succeed- 

 ing flabellar joint. On the smaller antennae of the blind Asellus cavaticus the olfactory rods are likewise much more 

 developed than in Asellus aquaticus provided with eyes. 



The olfactory rods in the species of Orconectes are longer, according to Faxon, than in normal 

 Cambari and Astaci. Leydig, who was the first to examine these organs, claimed that the olfactory 

 rods in these Crustacea were more numerously developed than in the normally eyed forms, and 

 that this was a compensation for the loss of sight. Leydig counted on the outer flagellum of the 

 first pair of antennae thirty six joints. Of the olfactory rods — for they are rod-like rather than like 

 teeth (Riechzapfen, as Leydig calls them) — all occur on the distal end of the flagellum; two occur 

 on the anterior end of the fifteenth joint, while on the succeeding joint there is a third one. On 

 the distal ends of most of the succeeding joints there are four rods and in the middle three, or 

 seven in all, but towards the end of the flagellum the number of rods again diminishes to two and 

 finally to one rod, their greatest development being in the second third of the flagellum. 



A single olfactory rod which Leydig figures (Taf. Ill, fig. 27) consists of a thick-walled stalk 

 and a more delicate terminal tube, which is thicker than the stalk. The rod is perforate at the 

 thickened end by a narrow opeuing which connects with the hollow interior. The substance of 

 the interior of the stalk appears to contain coarse granules, that of the terminal part fine ones. 

 Although Leydig does not say so, it is evident, in the light of the more recent studies of Hauser 

 and Kraepeliu, that this hollow olfactory rod, perforate at the end, is fundamentally homologous 

 with the olfactory teeth of insects. Our own observations on the rods of Orconectes hamulatus con- 

 firm the accuracy of Leydig's description of those of 0. pelhicidus (see PI. XXII, figs. 6, 6a, 6b, 6c). 



Prof. R. Ramsay Wright (Amer. Naturalist, xviii, 272) examined these rods in Cambarus 

 propinquus,- finding the external branch of the first antenna composed of eighteen or nineteen 

 joints. The distal nine of these alone bear olfactory cones, and only five of them (the eleventh, 

 twelfth, thirteenth, fourteenth, and fifteenth) have the full number of eight on each joint; hence 

 he concludes with Leydig that the olfactory rods ''are present to a much greater number than in 

 allied forms possessed of sight." 



There however seems a greater variation than suspected in the number of joints to the outer 

 flagellum, and it is the greater length of the rods in the blind species which indicates greater 

 olfactory powers. In his " Revision," etc., Professor Faxon makes the following statement : 



I have examined several specimens of C. propinquus with reference to this point, and find that the number of 

 segments in the external flagellum of the antennule may be as high as thirty-five, fifteen or sixteen of which may 



* As regards the latest statements, we might quote Leydig, Ueber Amphipoden und Isopoden. Zeits. f. wiss. 

 Zool., xxx, Suppl., 1878. S. Fries, Mittheilungeu aus dem Gebiete der Dunkelfauna. Zool. Anzeiger, 1879. Fries 

 states: On the short upper antennse of Asellus cavaticus not only do the number of joints vary, but also that of the 

 olfactory rods. Sometimes only the two penultimate joints, sometimes three or four joints, are provided with these 

 rods; in one instance I counted six such rods. Eougemont records five to seven long olfactory rods. The rods 

 themselves are in A, cavaticus very considerably longer than those of A. aquaticus. 



