Somatic chromosomes of higher diptera

Summary

The chromosome complements of eighty brain cells ofHylemya antiqua have been studied. The eighty cells were found in thirty-three larvae. Total complement length (TCL) is not randomly distributed among the larvae. Because there is an inverse correlation between chromosome length and width, it appears that in the cells studied the different chromosome lengths are partly expressions of different stages of metaphase contraction. It is suggested that synchronous division of cells still occurs in late larvae.

The length of each chromosome arm is highly correlated with that of every other arm. It is possible that the correlations are complete but that inadequate technique causes the departures from completeness which are observed. The chromosome lengths are corrected slightly for distortions, but the corrections make very little difference in the correlation coefficients. There is a high value for the correlation between the correlation of two arm lengths and the sum of the two arm lengths. This is to be expected if the perfect correlation between all arm lengths is being obscured by errors of drawing and measurement.

The autosomal arms have very similar coefficients of variation. The arm ratios (length of long arm divided by short arm) are not correlated with TCL or with each other, and arm ratio is randomly distributed among the larvae. The sex chromosomes have a smaller coefficient of variation than the autosomes, so that they are relatively large in small cells and relatively small in large cells.

Twenty-two cells inHylemya fugax were measured. The autosomes also showed a high correlation between arm lengths. An entirely heterochromatic autosomal arm showed the same phenomenon of a low coefficient of variation which was shown by the heteropycnotic sex chromosomes inH. antiqua. The low variability of heterochromatic regions accompanied by an apparently non-random distribution of the TCL may produce an erroneous picture of the species complement when dealing with small numbers.

It is suggested that for simplicity in using cytological observations of this sort for taxonomic purposes, the technique of measuring the percent TCL of a chromosome plus its arm ratio be replaced by the percent TCL of each arm plus the average length difference between the arms of each chromosome pair in units of percent TCL.