In the coming year we expect to give this problem further consideration.

Onion and tobacco seed constitute the bulk of the seed received for separation. Both of these crops are grown extensively in the Connecticut valley. The object of separation is, of course, to discard the small seed and chaff, leaving only the large and heavy seed for planting. The percentage of the seed discarded varies somewhat. This would naturally be expected, as the seed comes from different dealers and varies in size and quality, although much of the tobacco seed is home grown.

In some of our seed separation work the amount discarded depends upon the percentage of germination of the sample. Certain growers make a practice, before purchasing in bulk, of obtaining samples, which are sent to the experiment station for testing. A certain percentage of the small seed is blown out, and the original sample, as well as the large and small seed are tested for germination. The results are then sent to the grower, and if satisfactory he purchases in bulk and requests that a certain percentage be blown out.

Small seeds, like tobacco, which are received in small quantities, are separated in glass tubes with bulbs of a special form, and in this form of separation, which is fully described in Bulletin No. 121 of this station, we make use of fifteen pounds' air pressure. The larger seeds, like onion, are separated by a special winnowing machine, given a constant speed by an electric motor. The machine is so arranged that a bushel or more may be separated at one time.

Careful germination tests were made again this year to prove the value of seed separation, and experiments in planting the different grades of separated seed were also made in co-operation with growers in different parts of the State.

The following tables give an outline of the seed work which has been done the past year:

TABLE 1. - Records of Seed Germination, 1908.

The average germination of the onion seed for 1908 is not as high as for either of the two preceding years, some very poor seed having been received. Onion seed was occasionally brought in for testing which had been left over from previous years and was too old for use. The corn and lettuce seeds received were remarkably good samples, six lots of the lettuce seed giving 100 per cent. of germination, a very unusual percentage.

Only 2.5 per cent. was discarded from the best sample of onion seed, 4.6 from the best tobacco seed and 9.3 per cent. from the best celery seed. From the poorest onion seed 29.7 per cent. was discarded, 26.9 per cent. from the poorest tobacco and 33% per cent. from some of the celery seed. The large percentage discarded, however, does not in all cases imply that the seeds were poor; on the other hand, as much as 33% per cent. was frequently discarded purposely, especially when the seed was to be used for experimental purposes, which was the case

with the celery. The average percentage of germination shown is good when one considers the large amount of poor seed which is yearly placed on the market.

Farmers who use large quantities of seed are naturally more particular about its quality than those who purchase small quantities. Tobacco and onion growers and market gardeners may be considered specialists, and the quality and source of the seed which they purchase are important items with them, making it a matter of good business to obtain the best seed from the most reliable firms. Since practically all tobacco men grow their own seed they have an opportunity to exercise great care in the selection of their seed plants, and onion growers purchase from reliable dealers; but the general farmer and the individual who plants only his vegetable and flower garden need only small quantities of seed, and often purchase the so-called "packets" (commission seed) from the local dealer. These seeds are often not true to name and are too frequently worthless as regards quality.

Mr. Edgar Brown' in charge of the United States Department of Agriculture Seed Laboratory, found that the average germination of 2,778 packets, including 26 kinds from 27 seedpacketing houses, was only 62.2 per cent.

The average germination of seeds from one firm was only 37.3 per cent., and from another 44.3 per cent.

The variation in germination of different packets of the same kind of seed from the same firm was in several cases more than 90 per cent. By 21 of the 27 seedsmen whose seed was tested, 200 lots of seed were put up which germinated 10 per cent. or less.

By 13 packeting houses 62 lots of seed were put up which entirely failed to germinate.

The average germination of the "commission" seeds tested was 25.7 per cent. lower than that of those sent out in the congressional seed distribution during the past six years.

It is quite evident from the results of Mr. Brown's investigations on seeds and there is no reason to believe that the results are different than would be obtained from any testingthat we are sadly in need of seed legislation. However, the large dealers in seeds who sell direct to the consumer are not,

1 Bulletin No. 131, Bureau Plant Industry, United States Department Agriculture, 1908.

as a rule, unscrupulous, most of them testing their own seed, as they do not care to injure their business by unloading on the public seeds which are worthless. They do not, however, as a rule, guarantee purity or the percentage of germination.

Twelve purity tests of clover, alfalfa and mixed grass seed were made, the highest percentage being 99.4, the lowest 85.3. The impurities found in our seeds are largely weed seeds of a noxious nature. One farmer informed us that after seeding down his large piece of land it required considerable time and expense to pull up the dock which had established itself in his field as a result of using seed which contained seeds of this noxious weed. There is constant complaint in regard to the impurity of seeds, and a considerable burden is imposed on the farmer by his being obliged to waste time and money in exterminating the noxious weeds introduced in his grass seed. It should be stated, however, that all weeds found in newly seeded land do not necessarily come in the grass seed. The past summer many lawns planted with the best seed obtainable produced nothing but pigweed, crab grass, etc., for the soil moisture, on account of the drought, has been insufficient to start grass and clover, while it is one of the characteristics of many weeds that they possess a wide range of adaptability, and will survive under conditions which would be disastrous to grass and clover.

Seed to be tested or separated should be sent by mail or express to G. E. Stone, Massachusetts Agricultural Experiment Station. The work is done gratuitously by the station for people living in the State, but the postage or express charges should be paid by the person sending the samples.

REPORT OF THE ENTOMOLOGISTS.

C. H. FERNALD; H. T. FERNALD; J. N. SUMMERS.

During 1908 the entomological work of the experiment station has progressed along the usual lines. The correspondence has been extensive, requiring much time and often a considerable amount of study before all the inquiries received could be satisfactorily answered. The year has been a favorable one for the increase to destructive numbers of many different kinds of pests, and this condition has been reflected in the letters received.

The experimental work of the division has been the continuation of investigations previously begun, these requiring so much time as to practically prevent taking up additional lines of research. Tests of the resistance of muskmelons under glass to the effects of hydrocyanic acid gas have been continued, but have not as yet been completed. A determination of the best methods for the control of the cabbage maggot was started in the spring, but, as was the case the preceding year, almost no maggots appeared either in the experimental portion or elsewhere in the field, so that no results of any great value could be obtained. This, though disappointing, only signifies that the experiments must be continued until the maggots become sufficiently abundant to give real tests as to the value of the different treatments; and as these must be begun before the maggots appear, it is very possible that the work may require repetition for several years before any data can be obtained.

Observations on the dates of the appearance of the young of the more common and injurious scales have also been continued, adding to our knowledge of this subject and making it possible to more nearly set the date limits within which spraying for these insects must be done in order to be effective; but these observations have not thus far covered a sufficiently long term