612 Gro. W. TANNREUTHER, 
generations a few winged forms appear. There is, however, con- 
siderable variation in the number of winged forms in the second. 
generation of M. salicicola from a given stem mother; as observed 
in many instances, ninety-five per cent. of the offspring may become 
winged. These winged individuals go from the original host to 
another of the same species and start a new colony. In this way 
the aphids become more or less evenly distributed in a particular: 
locality. The succeeding parthenogenetic generations agree with 
the first and second in their habits and structure. 
The conditions of food and temperature are a very important 
feature in aphid development, though they influence only in an 
indirect way the appearance of the sexual forms. Starting with a 
given stem mother, favorable conditions promote rapid growth and 
hasten the reproduction of the succeeding parthenogenetic generations, 
while unfavorable conditions retard growth and lengthen the time 
for any given generation. 
From a series of observations made in the field and experiments 
in the green-house, in order to determine the time period for each 
succeeding generation and the number of parthenogenetic generations 
that intervened before the appearance of the sexual forms, it was 
found that external conditions, whether severe or normal, would not 
bring about the production of the sexual generation before a definite 
number of parthenogenetic generations intervened. Abundance or 
scarcity of food is not a factor in determining the sex in case of 
the aphids. This is shown beyond doubt in the presexual or last 
parthenogenetic generation where different individuals produce either 
all males or sexual females irrespective of external conditions. 
The minimum period of existence for any given generation in 
favorable conditions is fifteen days. In unfavorable conditions the 
maximum period is thirty-five days. If conditions are normal 
throughout the summer season the average time required for the 
completion of each parthenogenetic generation is about twenty to 
twenty-five days. This irregularity for the appearance of succeeding 
generations is due to food and temperature. 
In a number of instances where the host was in an abnormal 
condition, the required number of parthenogenetic generations (ne- 
cessary before the appearance of the sexual female) was not produced 
until the middle of November and very few winter eggs were de- 
posited. In more severe conditions no sexual females were produced 
at all. Normally, the sexual females appear about the middle of 
