28o 
B. W. WELLS 
The Sheath 
The results of this study are in accordance with Prillieux's (6) and 
Petri's (4) observations that an organic membrane is deposited by the 
cytoplasm around the setal proboscis (figs. 56, 6, 8, 10). I found as did 
Petri in the grape Phylloxera root gall that this membrane is composed of 
calcium pectate, but was unable as he was to demonstrate a layer of tannic 
substance on its exterior. In Petri's material the deposit was irregularly 
laid down or, as he described it, appeared "warty," while in mine it was 
deposited with remarkable uniformity as to thickness. This sheath is open 
at its end, thus making possible the direct movement of cellular substances 
into the end of the setal proboscis. This opening can be demonstrated 
only with the oil immersion objective in favorable sections (fig. 5&), for in 
many instances the end of the tubular sheath is surrounded by a dense 
mass of cytoplasmic granules. The nature of this mass I was unable to 
determine. 
Figure 8 illustrates a case in which the insect partially withdrew the 
setae, then plunged them in again in a different direction. Figure 10 
shows an extremely rare condition in which the nymph has withdrawn 
the mouth parts completely and twice reinserted them. The setae were 
found as shown, broken off and sticking in the shortest of the three sheaths. 
The middle sheath is distinctly abnormal. This section shows a condition 
commonly observed, viz., the projection of the sheath structure beyond the 
surface of the outer cells. 
Examination of half -grown galls and of older ones has shown but one 
sheath present, indicating that but one insertion of the proboscis takes 
place, the insect remaining passive and maintaining a stable position for 
relatively long periods of time. Rosen (7) found the same to be true in the 
grape Phylloxera leaf gall. Rosen's figure 4 shows a perfect sheath which 
he inadvertently misinterpreted as "proboscis." 
Since in all studies to date of Hemipterous insect cecidiogenesis, an 
organic sheath has been found surrounding the setal mouth parts, we have 
reason to believe that the phenomenon is a general one for that group of 
zoocecidia. 
Multinucleate Cells 
Multinucleate or giant cells are not unknown in gall tissues, and even 
are not absent from normal tissues according to Beer and Arber (i), who 
have found the phenomenon in the stems of 50 dicots and 17 monocots. 
Prankerd's (5) studies of the occurrence of multinucleate cells in normal 
tissue are of interest in this connection. He believes they are mostly formed 
by amitosis. He finds in many instances that a wall is formed later by the 
two protoplasts resulting from a single amitotic division. Kiister (3) 
presents a number of records of cecidial giant cells in many of which amitosis 
is reported as the mode of origin of the nuclei. In one instance, the Erineum 
