HIl'POLYTE VARIANS. 645 



tation of one eye produces little efllect. In the dim light or 

 towards evenings prawns thus treated often become beauti- 

 fully nocturnal, and recover quite symmetrically when 

 brought into light. Amputation of both eyes causes, in a 

 large proportion of cases, in brown specimens, a sudden 

 appearance of the green colour, and, if the specimens be then 

 put in the dark, a rapid transition to the nocturnal colour 

 (Table XI, p. 689, Specimen G). In speaking of periodicity 

 it was pointed out that amputated specimens nocturne and 

 recover, though somewhat irregularly, and we can now bring 

 additional evidence to confirm this. The preliminary green 

 effect which often followed all our methods appears due in 

 part to shock ; more largely, howevei", to suddenly cutting off 

 the light by way of the eyes (some, however, became reddish). 

 The most complete evidence of the behaviour of prawns 

 when amputated or " nitrated " in comparison with normal 

 control specimens is given in Table XI, pp. 688—691. Thus 

 G, originally a dark brown form (August 19th, 12.45 p.m.), 

 became greenish after removal of the eyes, and by 3.50 had 

 become nocturnal. It was then placed in a white porcelain 

 dish and exposed to daylight in the open till 7 p.m., aftei-- 

 wards to the incandescent light. By 9.30 both G and E 

 (a specimen whose cornea had been treated with silver 

 nitrate) had become semi-nocturnal, while the normal control 

 specimen (F) was now a full nocturne ; that is, the speci- 

 mens whose eyes had been operated upon had become 

 nocturnal, and then in part recovered before the time of 

 nocturning of uninjured specimens. G recovered fully by 

 9.50, E in two minutes, and both were now put with F in 

 a muslined jar. At 10.40 p.m. G and E were again 

 nocturnalish, F nearly full nocturne. 



Another example is C, which appears to show in addition 

 that amputated specimens respond more slowly to change of 

 light-intensity than do normal ones. The specimen operated 

 upon (C), and a control (I), were placed in a vessel beneath 

 a Landolt light-filter arranged to transmit red light. C 

 originally put into red light at 3.50, had, together with the 



