When a certain red-flowering plant is self-fertilized, genetic theory indicates that the plants developed from the resulting seed should be in the ratio of 3 red-flowering plants to 1 white-flowering plant. If a random sample of 100 such seeds is collected and 68 produce red-flowering plants, 29 produce white-flowering plants, and 3 do not germinate, do these results agree with the theory? Use a multinomial chi-square test with a = 0.01. What assumption must be made about the nongerminating seeds for this to be a valid test?