Some sources of error in the determination of chromosome length

Abstract

a) Three methods of measurement of straightened rye chromosomes were compared in respect to their accuracy. 1. Direct measurement with an ocular screw micrometer, using a 100 x oil immersion lens. 2. Measurement in a photograph taken with a 100 x oil immersion lens. 3. Measurement in a photograph taken with a 40 x dry lens. 1. and 2. were found to be of equal accuracy, giving a coefficient of variability of approximately 0.6%. 3. was less accurate, giving a coefficient of variability of approximately 2.2%. For rye chromosomes this is not serious, for smaller chromosomes it can become disturbing. Replication of measurement in the same photograph will then be useful; between photograph variability is small. b) The effect of squashing on chromosome length and on arm length ratio was studied for rye chromosomes. On the average squashing was found to increase chromosome length with several percent, arm length ratio by a few percent, both effects being statistically significant. In individual cases much more extreme increase of length and ratio was found. The effect is smaller than the variability due to other causes, even if not including the variability due to inaccuracy of the method of measurement. c) The effects of different pretreatments on chromosome length and arm length ratio in rye were compared. No significant difference regarding arm length ratio could be found between oxyquinoline, bromonaphthalene and colchicine in the concentrations used, although there are indications that such differences may exist but are masked by the large variability due to other causes.