10 BULLETI^s^ 276^ U. S. DEPARTMENT OF AGRICULTURE. 
labeled 31. pisi or its equivalent, and in no case has the true pisi 
been found. On the other hand, Dr. Edith M. Patch had no diffi- 
culty in getting this insect to hve contentedly on shepherd' s-purse. 
Theobald (11) lists shepherd' s-purse (Bursa hursa-pastoris) as a 
host, but later in his paper says: '^Colonies now and then occur on 
the shepherd' s-purse, but I have never known them to survive any 
length of time." The species reported for lettuce as M. pisi or a 
variety of that species by Sanderson and others is an entirely different 
plant-louse. Doubtless the species collected on nettle (Urtica) and 
referred to this species by Oestlund is something else, and the same 
can be said of the records of this aphidid from beet. 
In 1900 Dr. Chittenden reported tests made by ^Ir. Theodore 
Pergande to colonize this insect on the followmg hosts, but with 
negative results in each case: Sonchus asper, dandehon, shepherd's- 
purse, Sisymbrium officinale, and dock. 
Dr. Edith M. Patch (1911) has reported a series of insectary host- 
plant tests for M. pisi, which may be briefly summarized as foUows: 
Transfers from peas (Pisum sativum) to potato (Solanum tuberosum) , 
to barley, wheat, oats, purslane {Portulaca ohracea), beets, and 
squash were wholly negative; from peas to red clover partially 
positive, and from peas to shepherd' s-purse (Bursa bursa-pastoris) 
positive. 
During the late summer of 1911 ^Ir. C. W. Creel, of this bureau, 
and the writer conducted a number of transfer experiments, with 
the following results: From red clover to cowpeas, cultivated buck- 
wheat, wild buckwheat (Tiniaria cristata) , wild mornmg-glory (Con- 
volvulus arvensis) , fleabane (Leptilon canadense) , pepper-grass (Lepi- 
dium sp.), wheat, alfalfa, yellow sweet clover (Melilotus officinalis), 
and cinquefoil (Potentilla sp.), the results were negative; from red 
clover to soy beans were partially positive and indicated that the 
insect might survive and reproduce on young tender plants; from 
red clover to red clover, garden peas, and white sweet clover (2Ieli- 
lotus alba) the insect transferred readily and fed and reproduced con- 
tentedly. In Chicago, 111., we found it breeding very abundantly on 
tender succulent shoots of 2Ielilotus alba growing under greenhouse 
benches. 
Theobald (12) attempted to colonize the species on willow, rasp- 
berry, clematis, clover, and Lathyrus, but was successful with only 
the two last-named plants. 
!Mi\ C. E. Sanborn (1904) reports this aphidid from rose, but in a 
recent letter he writes: " ^lacrosiphum pisi has been correctly 
recorded as being taken on rose. I doubt, however, if rose should 
be considered as a regular food plant of this insect." We have 
repeatedly attempted to colonize pisi on rose, but without success, 
and there seems to be no reasonable question but that the specimen 
collected by Sanborn on rose was a stray migrant. 
