No it is not. Pumpkin/squash seeds float. There are a few others that float too but the majority of plants that you would wash the fruit away on sink.

I saw this post last week and had been wondering how to tell good seeds from bad on my vining plants. I wanted to save the seeds from a watermelon plant of mine, and tried the float test. Many more floated than sank, so I kept only the sinkers? Any good way to tell good seeds from Squash, Cucumber and watermelon?

With squash and cucmber I feel each individual seed - flat ones are not viable, plump ones are.

I have used the float test for Solanaceae (tomatoes, peppers, eggplant, tomatillo, etc.) and for most cucurbits (melon, cucumber, gherkin, etc.). The procedure must be performed on wet seed only, either directly after removal, or immediately following fermentation. Once seeds have dried, however, the test may adversely affect viability.

The float test works very well for eggplant, tomatillo, ground cherry, and Solanaceae exotics with a similar fruit structure. Just add the fruits (or thin slices, for eggplant) to a blender full of water, and pulse at the lowest speed until seeds have broken loose. The seeds are very hard, and few will be harmed; the good seeds will sink immediately. It is best to process a small quantity at a time, however... too much, and the thick layer of floating debris will trap many viable seeds. Whipping the debris layer vigorously with a wire whisk will release some of the good seeds, which will then sink.

Other seeds (such as peppers) tend to hold on to air bubbles, which must be knocked loose to get accurate results. Peppers are especially challenging, and must be processed quickly, before the seeds dry. Placing the seeds in a tall container full of water, I use a wire whisk (again vigorously) to dislodge bubbles & pieces of clinging membranes... this also cleans the seed. A blender can also be used (gently) but will sometimes make more bubbles than it knocks loose. I strongly advise rapid processing & drying if the float method is used for pepper seed.

For cukes, West Indian gherkin and Mexican sour gherkin (and probably kiwano), a day or two of fermentation (preferably in their own juice) is helpful, to more easily destroy the gel sacks surrounding the seeds. Stir & inspect the seeds at least twice a day, and process immediately when the gel sacks begin to break loose. Again, beating the floating debris with a whisk afterward will help dislodge any stubborn gel sacks that were not destroyed by fermentation. I have had good luck with brief post-fermentation processing in a blender for the smaller seeds, which are seldom harmed; but cukes & melons (and any seed that size or larger) could suffer substantial damage. Use of a dedicated blender or processor for seed processing (with plastic blades, or metal blades ground dull) will minimize the damage.

As mentioned by Oldherb, the float method is not reliable for most squash. I used it this year on acorn squash seed & many seeds did sink; but I believe some viable seed was also lost, due to stubborn air bubbles trapped in the membranes. Watermelon seed was difficult for the same reason. I believe that the float test for viability can be effective for most cucurbits... but only if some method (either fermentation or mechanical) is used first to dislodge the membrane surrounding the seed.

A variation of the fermentation method can be used for squash, where a little water is added to the seeds, and they are allowed a short fermentation/soaking period (I believe it was 24 hours or less). After soaking, they are beaten (carefully) or squeezed to free them from any remaining debris. The viable seeds, having absorbed water, will sink; the floaters can be skimmed or carefully poured off. The seeds must then be strained _immediately_, and dried rapidly under fans (or preferably, in a dehydrator). Glen Drowns of Sand Hill Preservation uses this method quite successfully... but personally, I would recommend it only to experienced seed savers, since there is great risk of sprouting the seed if done incorrectly. It is often far easier to remove the few bad seeds visually.

So in summary, the float test can be very effective for tomato, eggplant, most small-fruited Solanaceae, cucumber, melon, and small-seeded cucurbits like West India gherkin & Mexican sour gherkin. For peppers, watermelon & the smaller-seeded squash, there may be some loss of good seed... but the quality of the seed that sinks will be very high. If the gel sacks of cucurbits are not removed, the float test will not be effective, and should not be used. For the larger-seeded squashes, I believe that hand selection of fat seed remains the best method for most gardeners. I will continue to use the float method where possible, since it yields high-quality seed... but for small quantities of the more difficult seed, it may be impractical.

Just want to add to the previous post... yesterday I cleaned seeds of "Tromboncino" squash, which is C. moschata. There were a lot of "duds", so I tried to sort them using the float method. I removed as much of the clinging flesh as I could by rubbing the seeds on a screen under running water, then put them in a plastic pail with a lot of water, and beat the *! out of them with a wire whisk.

The whisk dislodged most of the flesh still remaining, and many of the seeds sank quickly to the bottom. I skimmed the floaters off & examined them; there were none that I would have kept. Then I examined the seeds that sank; there were some that were discolored or deformed, but most appeared healthy. I'll do a germination test after the seeds have dried, to see how much is viable, and post a follow-up.

But for now, the float method seems to work for separating C. moschata squash seed without fermentation, as long as most of the clinging flesh has been mechanically removed. Some minor culling of the seed that sinks will still be necessary.