NEW YORK EXPERIMENT STATION BULLETINS. 349 



hundred and eighty-five squashes and gourd 'flowers of some fifty varie- 

 ties were individually pollinated. One hundred and sixty-three of these 

 did not produce fruit. The remaining twenty-two carried fruits to matur- 

 ity, but in every case these seeds were thin and worthless. These 22 

 fruits represented 13 bush summer squashes of various kinds, five small 

 ornamental gourds, and four crosses between bush squashes and gourds. 

 In cross-pollinations made during the same time and in the same manner, 

 a large part of the crosses were successful, indicating that the failure of 

 the individual crosses was due to the inability or the pollen to fertilize 

 the ovules rather to incidental methods of operation. The experiment indi- 

 cates that pollen of squashes which can not produce fertile seeds may still 

 cause the development of the fruit. The influence of pollen is well 

 attested in other instances, but it is not impossible that squashes may 

 sometimes develop without any pollination whatever. At any rate we have 

 found this to be the case in some other cucurbits, and it is a point upon 

 which we are still working, and concerning which we have much data. 



This impotency of individual pollen is a matter of immense importance 

 to originators of varieties. It is commonly held that the best way in 

 which to "fix" or render permanent new varieties, so that they will repro- 

 duce themselves by seedage, is to in-breed or close-pollinate them, but the 

 above trials indicate that this is impossible or practically so in pumpkins 

 and squashes. An instance in my own experience is suggestive. From 

 stock which was crossed in 1887, I obtained in 1889 one squash of great 

 excellence. It appeared to combine more good qualities than any squash 

 of its type that I have ever seen. To procure as many plants of it as 

 possible, in order to cross and fix it, I planted all the seeds from the 

 best fruit in the spring of 1890. These seeds, all from one squash, pro- 

 duced 110 distinct varieties, and only one plant was like the parent! The 

 only thing to do was to pollinate the flowers of this one plant with pollen 

 from itself, but it soon became evident that all of these individual pollin- 

 ations would fail. It was then necessary, late in the season, to pollinate the 

 remaining flowers from some other plant which bore fruit the nearest like 

 the one under experiment. Fortunately, two or three other plants bore 

 similar fruits, and by the use of their pollen two good plants were 

 obtained. 



It appears, therefore, that in squashes and pumpkins the pollen is 

 impotent upon pistils on the same plant, and that true in-breeding does 

 not occur in them. The experiment will be extended to all varieties. 



4». Do cucumbers spoil musk-melons? — If any dogma finds general 

 acceptance among horticulturists, it is the opinion that musk-melons are 

 rendered insipid and worthless by cucumbers growing in their vicinity. 

 Most growers suppose that this influence is immediate, but a few hold that 

 it appears only in the offspring of supposed crosses between the two spe- 

 cies. Several years ago my observations led me to doubt this influence, 

 but definite experiments were not undertaken until last winter, when a 

 house of forced melons and cucumbers gave a good opportunity to make 

 cross-pollinations. In these trials we failed to produce melons when the 

 flowers were pollinated either by the common white spine or the English 

 forcing cucumbers. 



Last summer the work was undertaken in the garden under the best of 

 opportunities. Ninety-seven musk-melon flowers of various varie- 

 ties were pollinated by cucumber pollen of many kinds. No fruits 



