TIMOTHY H. GOLDSMITH 791 



distinct (22, 06, 8, 13, 67), and it is generally stated that there is a 

 nerve fiber Ironi each retinnla cell. In the lour other orders of in- 

 sects [Odonata (22), Orthoptera (13), Lepidoptera (13), and Hy- 

 nienoptera (21) ] that have been examined with the electron micro- 

 scope, however, the rhabdom is a Insed structme, and it is not at all 

 certain that an excitation arising in one of the rhabdomeres would 

 be transmitted exclusively through the corresponding retinula cell. 

 Furthermore, counts of ganglion cells in the lamina ganglionaris of 

 the honeybee indicate probably no more than one second-order unit 

 for each ommatidium (43), an observation which suggests that even 

 if the individual retinula cells are capable of responding indepen- 

 dently, their fibers converge extensively at the first synapse. It seems 

 not to be the case, however, that all the receptor axons from a given 

 ommatidium synapse with the same ganglion cell, a relation in- 

 dicating that the receptive fields of the ganglion cells may be larger 

 than single ommatidia (eg. 25) . 



The ommatidia of Limulus respond as units (40) . However, the 

 presence of the eccentric cell — a neuron which appears to have 

 migrated out into the layer of sense cells (41) — may make the horse- 

 shoe crab something of a special case. There are no comparable 

 electrophysiological observations of individual ommatidia of insects. 



Hecht and Wolf (27) found that the visual acuity of the honeybee, 

 as determined by the optomotor response, corresponds to the interom- 

 matidial angle. Similar results have been reported for other insects 

 by other workers. As the visual fields of adjacent ommatidia are now 

 reported to overlap extensively (53) , this correlation may be a coin- 

 cidental result of sharpening of contrast based on such a neural 

 mechanism as the inhibitory interactions between ommatidia described 

 for Limuhis (26) . 



There is good evidence that the receptors for color, whether they 

 Ije single ommatidia, retinula cells, or groups of retinula cells, may 

 not be distributed uniformly throughout the compound eye. Ac- 

 cording to Rokohl (45), only the dorsal-posterior ommatidia of the 

 eye of the back-swimmer (Notonecta) subserve color vision; the rest 

 of the eye is color-blind. Similarly, Walther (57) finds that a high 

 sensitivity to ultraviolet light is limited to the dorsal half of the eye 

 of the cockroach (Periplaueta) , whereas a green-sensitive receptor is 

 found in all parts of the eye. In the honeybee the spectral sensitivity 

 depends to some extent on the position of the electrode in the eye, 

 but the distribution of receptors seems to be different from that ob- 

 served in the cockroach and has not been adequately mapped (20) . 



