HERPETOMONAS ELMASSIANI. 329 



one plant to another. But since there are several reasons for 

 suspecting it to be the carrier, a study of its flagellates was 

 undertaken. All along the coast of America, from Paraguay in 

 the south to New Jersey in the north, wherever the flagellate 

 has been found in the latex of the milkweeds some species of the 

 genus Oncopeltus has been found feeding on the infected plants. 

 These insects have always been more characteristic of the 

 particular plants harboring the flagellates than any other type 

 of insect. Just north of New Jersey, where no flagellates are 

 known in the plants, another insect genus replaces Oncopeltus. 

 The specimens of this which have been examined thus far have 

 been entirely negative for any flagellate, though infected with 

 Sporozoa both intracellularly and extracellularly in the salivary 

 glands. This coincidence of the range of the milkweed flagellate 

 with the range of Oncopeltus fasciatus (Dall.) suggests that the 

 presence of the flagellate is dependent on the presence of the 

 insect. Moreover Oncopeltus feeds characteristically on the pods 

 and flowers of the plant, and since plants bearing seed are the 

 ones to which the infection is practically limited it seems even 

 more probable that Oncopeltus is the host. Another bit of 

 circumstantial evidence is gained from the study of the mor- 

 phology of the flagellates from the insect and from the plant. 

 The plant flagellates are characterized by a twist in their ribbon- 

 like bodies. This is rare in insect flagellates, but is found in the 

 case of the parasites of Oncopeltus. The insect feeds on latex. 

 The infected plants have swarms of flagellates in their latex in 

 every portion of the upper parts of the plants. Thus the insect 

 could easily become infected from his feeding. In the insect the 

 histological studies about to be described have shown swarms of 

 flagellates in definite lobes of the thoracic salivary gland. The 

 secretion of this gland is led by a simple duct directly to the 

 mouth parts during the process of feeding, so that there should 

 be no great mechanical difficulty in the transfer of the insect's 

 flagellates to the interior of the latex ducts of the plant. 



It is natural then that in spite of the difficulties which have 

 been experienced up to this time in attempting to obtain plant 

 infections from the insects in regions where plant infections are 

 not known in the field, efforts to establish the relation of the 



