314 Mr. G. A. K. Marshall mi 



Mantis \l., Vhyllocrania in&ignis, on the Lower Umko- 

 niaas River. The evidence that starvation for twelve and 

 fourteen days respectively does not produce the symptoms 

 observed in Experiments III., IV., V., VII., and IX. 



Mantis VII., male, in pupal stage, probably Poly- 

 sjnlota caffra, at Malvern. The Pierines T. achinc and 

 B. scvcrina freely eaten. Of the Acrffias, two cnccdon eaten 

 apparently freely, and one after an interval, others re- 

 fused : one screna eaten after a time : of four pii/ictatlssima 

 only one partly eaten. The Mantis then refused all food, 

 became weak, and one eye was affected. He was unable 

 to throw off the pupal skin properly. In Experiment VI. 

 two individuals of another species performed this change 

 of skin after ton and seven days of starvation. 



Mantis VIII., female, Pscudocrcohotra ivalilhcrgi, at 

 Malvern. Only offered Acra3as. Nine pwiictufissinia always 

 refused Avith or without trial ; cncedon eaten freely several 

 times, refused once, and partly eaten once ; two cahira eaten 

 freely ; serena eaten freely or after trial ; neohulc eaten 

 after two days' interval. 



In spite of this diet the Mantis remained apparently 

 healthy, September 2G to October 9, ]8!)7, when the 

 experiment came to an end. 



Mantis IX., female, SpJiodromantis lineola, at Salisbury. 

 It was intended to offer this individual a purely Acrma 

 diet, but she ate her mate on the third day after their 

 capture in cojrnld. She was chiefly fed upon Acriea 

 ccddarcna, which she ate sometimes freely, sometimes 

 after an interval and after trials : at other times she 

 refused it with or without trial ; one induna was eaten after 

 a day's interval ; one ncolmle was untouched ; one natalica 

 was partly eaten ; two halali were rejected after trial, one 

 without. After a fortnight of this diet the Mantis became 

 Aveak, and her left eye was affected : a day later she began 

 to nibble off the end of one of her fore tarsi : two days 

 later she oviposited, but the egg cocoon was only half its 

 usual size (eggs infertile). After the first signs of weak- 

 ness the Mantis ate only a small part of three Acra'as out of 

 eight offeretl her during six days. She continued to nibble 

 at her tarsi, lost power over her movements, and died 

 after twenty days of ca])tivity. 



Mantis X., female, Sphodromantis lineola, Salisbury. 

 This individual was the subject of a control experiment, 

 being fed solely upon several species of the following groups 



