208 



mens of the same scale on Blackberry collected by Mr. T. G. Chamber- 

 liu at Santa Barbara, Cal. In Bulletin No. 5 of this Division (p. 25) 

 other specimens of this species are recorded as having been reared from 

 a Mytilaspis upon an undetermined species of Dycaste from Japan, 

 which we received in 1874 from the late Dr. George Thurber, of the 

 American Agriculturist. The fact that this species is now found to 

 attack the Eed Scale (and we consider Mr. Coquillett's observations to 

 indicate more than a mere probability) is interesting and important, 

 since but one true i)arasite has jn-eviously been recorded from this in- 

 sect, viz, the so-called Coccophagm citriims Craw, which does not seem 

 to be multiplying rapidly. 



PARASITISM IN BEES OF THE GENUS STELIS. 



That the Apid genus Stelis develops in the cells of the allied genus 

 Osmia has been knowm for some time, but the exact nature of the j)ar- 

 asitism, and more especially when and how the Osmia larva is destroyed 

 by the Stelis larva, have hitherto not been explained. In a recent 

 number of the Zoologischer Anzeiger (vol. xv, No. 383, February 1, 

 1892, j)p. 41-43), Mr. C. Verhoeff, of Bonn, Germany, summarizes the 

 results of a series of careful observations which throw a flood of light 

 on the subject. The species observed are Os)nia Jeucomelacna K. and 

 Stelis minnta Nyl. 



The species of Osmia construct cells in the interior of hollowed twigs 

 in the manner of Megachile and similar bees. At the bottom of the 

 cell the female Osmia first iDuts a layer of pollen which is to serve as 

 food for the nearly full-grown larva. Above this pollen, the bee com- 

 mences to store the cell with prepared bee-bread. At this moment the 

 female Stelis watches her opportunity to lay an ^gg in the Osmia cell, 

 the egg thus being always near the bottom (posterior end) of the food 

 mass. LTiuiware of the presence of the parasite egg, the Osmia female 

 continues her work, and, after nearly filling the cell, deposits her own 

 egg on the top (anterior end) of the food-mass. The cell is then closed 

 with a layer of macerated particles of plants and a second cell pre- 

 pared above the first. The Stelis larva hatches but little earlier than 

 that of the Osmia, and both larvie feed on the food -mass, the parasite 

 larva at the bottom, the host larva at the top. The latter remains sta- 

 tionary at the top and grows very slowly ; the parasite larva grows 

 more rapidly, and gradually works its Avay upward through the food- 

 mass, thus gradually approaching the Osmia larva. The crisis finally 

 comes; the Stelis larva encounters the Osmia larva^ — a short but deadly 

 combat ensues — the Osmia larva is easily overpowered and killed by 

 the much larger and stronger parasite, and its body is devoured by the 

 latter within one or two days. 



It is thus evident that Stelis furnishes another illustration of that 

 partial parasitism which I have sIkjwu to be the rule with the MeloidtB? 

 but differs in that the parent iutroduces her Qgg into the host cell in- 



