176 Life-Hutonj of the Larcli Chermes 



absence of Sistens larvae. It is noteworthy that Sexupara stages 

 developed, though in small proportions, from these Progrediens larvae. 

 In May, 1919, the observations were repeated with egg-clusters 

 obtained in the same locality. 



Total 425 258 683 



It is clear that variation exists from year to year in the proportion 

 of Progrediens and Sistens eggs produced by the adult Sistentes. How- 

 ever, there seems to be some system in this variation. 



First, a large proportion of the egg-clusters gave rise only to Pro- 

 grediens larvae (all in 1914, 4 out of 7 in 1919). Secondly, some clusters 

 produce a very large majority of Sistens larvae, in which case the eggs 

 hatching first in the cluster were Sistens, and not Progrediens larvae. 



Thirdly, in common with Mr Steven's observations, some clusters 

 produce a majority of Progrediens, and a minority of Sistens larvae, in 

 which case the first eggs hatching gave rise to Progrediens, and only the 

 later ones to Sistens larvae. 



As Mr Steven points out, observation extending over many years on 

 isolated trees is necessary to determine with exactitude whether the 

 Sistentes of the three origins produce constant proportions of the two 

 forms, but the above statistics may well point to such research having 

 a definite result, and perhaps even to the trend of that result. 



B. Progredientes. 



False proportions are easily obtained if egg-clusters, from which 

 some of the larvae have already hatched and wandered away, are Icept 

 under observation. 



From seven such clusters collected at Oxford on March 28th. 191 1, 

 17 Progrediens and 25 Sistens larvae hatched between that date and 

 May 18th. In June, 1913, five similar clusters produced 30 Progrediens, 

 and 30 Sistens larvae, between June 14th and June 24th. 



These represent the first Progrediens generation, in which the pro- 

 portions seem to be fairly equal. 



