1 1 8 
Psyche 
[March-June 
had a chromosome number of 22, three had one of 24, two had one 
of 25, and one had one of 26. Karyotypes of 22, 24 and 26 chromo- 
somes are shown in Figure 1. Because of the small size of 
the chromosomes, we could not determine the nature of the poly- 
morphism, but Robertsonian changes do not appear to be involved. 
The range of chromosome numbers between 22 and 26 is probably 
continuous, with the lack of a count of 23 being due to sample error. 
This is the second report of intraspecific variation in chromosome 
numbers in Symphyta. Sanderson (1971) reports that, while Eng- 
lish collections of Tenthredo acerrima consistently had a haploid 
number of 18, Scottish collections had haploid numbers ranging 
from 18 to 21. There is thus a geographic basis to the T. acerrima 
polymorphism that remains to be searched for in J. integer. 
Chromosome numbers in the Apocrita range from four to 28, 
which range is found in the single family Formicidae (Crozier, 
1970). The low haploid number ( n — 9) (Mackay, 1955) of 
Cephas cinctus, combined with the fact that generally lower num- 
bers have been reported for Parasitica than for Aculeata, suggested 
that low numbers might be primitive in the Apocrita. However, the 
discovery that another cephid, Janus integer , has high chromosome 
numbers indicates that high numbers are as likely to be a primitive 
character in the Apocrita as low numbers. 
Karyotype frequencies are best considered using the genus-karyo- 
type concept (Crozier, 1970), in order to minimize the effects of 
different intensities of sampling between genera. Under this con- 
vention, each chromosome number is counted once each time it oc- 
curs in a genus, irrespective of how many species it occurs in. The 
genus-karyotypes known for the Symphyta are shown in figure 2, 
where it can be seen that Janus integer has the highest chromosome 
number of any sawfly. The distribution of genus-karyotypes is 
strongly skewed, suggesting that low numbers might be primitive 
for the Symphyta. The ranges in chromosome numbers of the fam- 
ilies Tenthredinidae and Cephidae are particularly striking. The 
gaps in these ranges might indicate polyploidy, but the karyotypes 
of too few species are known to give much support to this suggestion. 
Summary 
Eight males of Janus integer had haploid numbers of 22, 24, 25 
and 26; presumably a larger sample would have yielded counts of 
23 as well. The polymorphism does not appear to be Robertsonian 
in nature. The significance of these results to considerations of 
hymenopteran karyotypic evolution is discussed. 
