56 BULLETIN 903, IT. S. DEPARTMENT OF AGRICULTURE. 
been examined and its interior subjected for a few minutes to a tem- 
perature several degrees in excess of that obtaining in the trench 
immediately preceding the examination. 
Plate VII illustrates the details of the cages used for observing 
the phylloxeras on living roots. By means of the pulley and stand 
the cages were hauled up and set for examination. It is obvious that 
only young vines could be used for this work, as 9-inch pots were the 
largest used. The vines were planted in early spring, certain of the 
longer roots, drawn through holes cut in the saucer supporting the 
bottomless upper pot, being planted in the lower pots. Thus about 
4 inches of root between upper and lower pots were available for 
inoculation and observation. At the upper and lower ends of this 
visible portion the root passed through glass cylinders, and the inter- 
vening spaces between the root and cylinder were filled with sand 
and cotton (sand only was used in the lower cylinders) to prevent 
the escape of phylloxeras to the invisible portions of the roots, both 
above and below. For the viniferas and nonresistant American 
vines this, however, failed to answer the purpose in many cases. 
Out of 22 upper pots, which were examined several months after 
the exposed roots were inoculated and had suffered more or less 
severe infestation, 18 developed infestation on their roots, show- 
ing that phylloxeras had found their way up to the roots in the 
upper pots. Out of 36 lower pots liable to infestation on their 
roots by reason of the fact that the exposed portions of the roots 
above were infested, the roots in 9 showed no infestation or indi- 
cations of any previous infestation, whereas in 13 others infestation 
occurred which had resulted from larvae successfully penetrating the 
lower glass cylinders ; in the remaining 14 pots, infestation or signs of 
previous infestation occurred resulting from wanderers reaching the 
rootlets by penetrating cracks in the soil. In the case of the re- 
sistants, the cylinders of sand and cotton packed between roots and 
glass were effectual in preventing spread to the invisible portions 
of the roots. On these vines the infestation was always very slight, 
and the phylloxeras exhibited very little desire to travel. On a 
Champini (rupestris X candicans), on which the phylloxeras in- 
fested only the side rootlets, and which bore only a slight infesta- 
tion, wandering larvas entered the soil and infested the rootlets of 
one of the lower pots, but there was no penetration through the glass 
cylinders. 
As temperature is a factor of importance in the development 
of the phylloxera, the following comparisons (Table XII) of tem- 
peratures are noteworthy, taken (1) inside the cages containing 
living vines, (2) 2 feet below the soil surface, at a point in the 
laboratory vineyard a few feet distant from the trench containing 
the cages aforesaid, and (3) in the laboratory cellar: 
