10 BULLETIN 1076, U. S. DEPAETMENT OF AGEICULTURE. 



were all isolated before emergence and remained virgins throughout 

 their lives, so these figures are doubtless below the normal. 



FOOD PLANTS. 



Unless some food plant other than lotus is inhabited by this insect, 

 it is difficult to explain its presence at Kimberlin Heights. From 

 present knowledge of its habits it is inconceivable that it came with 

 the seeds, and no other lotus is known to exist within possible range 

 of flight of this moth. To determine whether the species is indigenous 

 or introduced, the writers have made plantings of lotus seed in several 

 isolated ponds many miles distant from this infestation. 



Because of the confusion in the literature between this species and 

 Pyrausta ainsliei. several food plants have been attributed to it which 

 manifestly are erroneous. The only natural food plants which have 

 so far been reliabh'^ ascertained are the j^ellow lotus, Neluvibo lutea, 

 and the Indian lotus, A^. nucifera^ the latter an introduced species. 



Smith {10, p. 525) , mentions having found these larvae in stems of 

 cat-tail flag, Typha latifoli-a. At Kimberlin Heights conditions were 

 ideal for such a transfer, beqause the lotus and cat-tail grew inter- 

 mingled in several places (PI. IV, A). In attempts to find where the 

 larvae went for the winter, practically every cat-tail jDlant in the 

 vicinity of the pond was thoroughly dissected and examined. With 

 the single exception of one larva found behind a leaf sheaf no trace 

 of attacks on this plant were found. 



In confinement in the laboratorj^ partly grown larvae, taken on 

 lotus, fed readily and completed their growth on leaves of smart- 

 weed {Polygonum fentisylvanicum) , buckwheat {Fagcpyrum fago- 

 pyr-um), and dock {Rumex crispus). In other series, larvae were 

 reared from egg to adult on the same plants but the authors have 

 never seen any indication that these plants are used as food by this 

 insect under natural conditions. Numerous aquatic and subaquatic 

 plants and a large number of the common wild plants and weeds 

 were offered to the larvae, but all were refused except those men- 

 tioned. It is noteworthy that the normal food plants of P. ainsliei 

 are Polygonaceae but that that species can not develop on lotus. 

 There is a suggestion here of some common ancestry for the two 

 species, with members of the Polygonaceae as their food plants, and 

 that P. penitalis^ having taken to lotus comparatively recently, has 

 not entirely lost its taste for the smartweed family. 



ENEMIES. 



In the course of its life several perils threaten the safety of this 

 insect. It does not seem to us that Welch's point (7^, p. 218). as to 

 the construction of the silken web being a special adaptation to its 



