About 1860 seedcotton cleaners with beater type cylinders began to be developed and by about 1910 the type of machine now called a cylinder cleaner had evolved (1). A number of spiked cylinder designs were used and a general configuration now found in most cleaners was developed. There is no published data showing the effect of the length and spacing of spikes or other cylinder design variables on cleaning performance. A test was conducted to determine the effect of doubling the spike population on a "standard" cylinder. The test was conducted with as-harvested and precleaned machine-stripped cotton at low and very high moisture levels. The standard cleaner had 12 rows of 1-7/8 inch long spikes around a 16 inch tip diameter cylinder. The spike population was doubled to 24 rows by adding 12 more rows of spikes at the midpoint between existing rows.

No differences were found in trash content of the seedcotton after cleaning due to increasing the spike population. Lint grade, staple length, micronaire, strength, and nonlint content were not significantly affected by doubling the number of rows of spikes. Increasing the number of spikes, however, appeared to slightly reduce fiber reflectance.

High moisture levels adversely affected trash extraction performance and reduced fiber grade and color values, but preserved fiber length and length uniformity.

The experimental cleaner was tested as an added unit at two positions within a conventional overhead cleaning system recommended for machine-stripped cotton. The position of the added unit in the overhead system was found to be highly important. The best overall system performance was obtained when the added unit was used as the first machine in the cleaning sequence. This arrangement produced lower foreign matter levels than did an arrangement where the added unit was positioned at the midpoint of the cleaning system.