M. M. METCALF ON THE EYES AND SUBNEURAL GLAND OF SALPA. 365 



cells in vertebrate eyes, which always show a similar thickening of their 

 cell walls. 



Professor Butschli says the horseshoe-shaped eye is secondary, being 

 derived from a " hiigelartig " eye, of which he gives a schematic figure. 

 His explanation of how the horseshoe-shaped eye is derived from the 

 "hiigelartig 1 ' form is so very neat that it predisposes one in its favor. 

 Let us, however, examine it. There is no eye among the solitary salpae 

 to which his description ("hiigelartig") could at all apply. Unless he 

 studied some very aberrant species which I have not seen, he must refer 

 to such an eye as is found in the chain form of Salpa runcinata-fusi- 

 formis or some of its nearest allies. In this group the larger dorsal eye 

 is either spindle-shaped or globular and answers roughly to his descrip- 

 tion. He says : " Die Hauptmasse dieses Hiigels [the eyej wird von 

 einer lichtempfindlichen Retina gebildet, deren Zellen der freien W61- 

 bung des Hiigels entsprechend, etwas radiar zu der Oberflache gestellt 

 sind. An seiner basalen Peripherie wird dieses hugelartige Auge 

 von einem ringformigen Giirtel von Pigmentzellen umzogen . . . Das 

 geschilderte einfache Auge empfangt seiner Nervenfasern direct von 

 unten aus dem Gehirn." The structure of this eye is not so simple as 

 this description says. In none of the eleven species I have studied do 

 the rod cells all have the same arrangement. In every case, as I have 

 shown, where the eye is globular or spindle-shaped, the rod cells of the 

 posterior portion of the eye have their innervated ends distal from 

 the ganglion, while in the anterior part of the eye this arrangement is 

 reversed. The pigment cells also show an arrangement different in 

 these two portions of the eye, and do not form what can be called a 

 simple girdle at its base. The rod cells, instead of receiving their inner- 

 vating fibers direct from the ganglion, receive them from a large optic 

 nerve which passes up over the dorsal surface of the posterior part of 

 the eye and then through to the ventral side of the anterior portion. 

 These three mistaken characters in his description are three of the four 

 points he emphasizes to show the primitive character of this form of 

 eye. The fourth point is its simple shape. 



Again, we would not naturally expect to find the primitive type of 

 salpa eye in the chain form of any species, since the chain individuals 

 are asexual, 1 and are derived from the solitary, sexual form by a peculiar 

 kind of budding. The chain individuals, then, being secondary, the pre- 



1 Professor Brooks has often emphasized this asexuality. In the present work, of 

 which my paper forms a part, he shows it with perfect conclusiveness. 



