Experiments with modified techniques for the determination of purity and viability of Bluegrass seed, Poa Pratensis L.
1. Determination of absolute purity in small seeded grasses with multiple florets, such as bluegrass, is possible experimentally but is a practical impossibility and when applied by different seed laboratories results in variations much greater than may be expected from homogeneous sub-samples of seed.
2. A close approximation of pure bluegrass seed may be obtained by a modified method of procedure which involves separation of unattached infertile florets or infertile spikelets from fertile florets by means of a uniform speed motor and fan and a vertical air blast separator of the Holland type. The only hand labor involved is that .of removing pieces of stem, grit, stones, weeds and other crop seeds from the heavy portion. A glass tube of the Kny-Scheerer type may be equally satisfactory, but breakage is too frequent.
3. Application of the chi-square test to four or more sub-samples drawn from the same lot and analyzed by the modified method shows that, in general, the percentages of heavy and pure seed fractions fall within the range of natural variability, thus indicating that the uniform speed motor and fan provide a reasonably constant air pressure.
4. Uniform performance of the motor and fan was also demonstrated by germination tests of the pure seed fractions from duplicate sub-samples of seven different lots in which no significant differences in germination of the replicates were obtained.
5. An additional evidence of uniform performance of the motor and fan was noted in the fairly uniform number of normal sprouts obtained in the inert fractions of samples having different weights per bushel.
6. Samples ranging in bushel weight from 16 to 27 pounds responded similarly to the same treatment thereby suggesting that weight per bushel for the samples studied was controlled more by the amount of empty florets present than by immature or light weight seed. Seed weighing 10 pounds required a stronger air blast than did heavier seed in order to remove infertile florets, although possibly a longer time at the standard opening might accomplish the same result.
7. Eight seed laboratories using the modified method of procedure, each with a similar sub-sample previously blown to aid in calibration of each blower, obtained purity percentages within the range of natural variability.
8. The adoption of the modified method of analyzing bluegrass for purity would involve the following: a. Installation of a standard control machine in a laboratory where frequent checks on samples could be made for all laboratories with similar equipment. b. Purchase of a scale and use of graduated vials for determining weight per bushel of small samples. c. Acceptance of the principle of approximate purity by a uniform method of procedure as a substitute for the ideal of absolute purity. d. Regular examination of pure seed remnants with a binocular or reflected light to check the reliability of the fan.
9. The use of a standard method of procedure based on the principle of separation of fertile and infertile florets by air pressure, as developed and maintained by a uniform speed motor, would solve one of the most difficult labeling and analytical problems now faced by the seed trade and seed analysts, respectively, and in addition provide uniform pure seed fractions for germination, which should automatically reduce some of the differences obtained in germination tests.
10. Germination tests of non-fresh seed of bluegrass representing 15 different lots each in quadruplicate indicate that sterile sand in petri dishes saturated with distilled water furnishes a slightly superior sub-stratum to filter paper moistened with distilled water or with .2% KN03 solution. Limited trials with copper trays equipped with wicks to maintain a constant supply of water indicate that they are equal to sand.
11. Total germination of bluegrass seed from 15 lots was significantly higher on sand than on filter papers at 10, 15 and 20 days after the tests were started. Mean germination at the end of 15 days on sand was practically equal to that on filter papers at the end of 28 days.