EMBRYOLOGY OF THE SEA LAMPREY 



131 



Table 3. — Specimens and number of eggs used in experi- 

 ments on development at constant temperatures 



(Ono group of lamprevs provided the eggs for experiments conducted at 45°, 

 55°. 7n°, 75°. and 80° and another group for those at 52.5°. 00°, and 77.5°. 

 Eggs from the lots that supplied materials for experiments at 50° and 65° 

 were reared at the one temperature only] 



was increased to 4 lioiirs at about 32 hours; later 

 samples (that is, after 2-3 days) were obtained 

 about 12 hours apart; and, finally in the lonpjer 

 experiments sampling was daily. Sampling after 

 stage 14 appeared was biased because the pro- 

 larvae had to be pursued as they became older; 

 thus the samples were non-representative in com- 

 parison with the random samples for all earlier 

 stages. In the main the bias led to over-repre- 

 sentation of live specimens. 



Several criteria were used to separate live and 

 dead embryos. The most obvious indication of 

 deatli was disintcErration of the embryo- a sepa- 

 ration of cells and subsequent filling of the intra- 

 membrane volume with loosely arranged cells, 

 within an intact membrane. Furthermore, the 

 fertilization membrane became translucent «itli 

 a cloudy cast in contrast to the transparency of 

 a living membrane. This difference was most 

 apparent in later development, stages 7-14; tiie 

 earlier stages (1-6) became vacuolated as they 

 underwent changes after death. Eventually, the 

 membranes of eggs in these earlier stages also be- 

 came cloudy. Another indicator, particularly in 

 early development, was a change of color from tlie 

 creamy white of normal eggs to a brownish tan, 

 accompanied by a fuzzy appearance of the surface. 

 Dying embryos of stages 9-11 possessed a])preci- 

 ably widened blastopores, which in some e.xteiided 

 across the entire diameter of the embryo. 



The basic data on tlie several temperature 

 experiments are given in the records of number of 

 dead embryos and number of living embryos in 

 various stages of development. In conjunction 

 with these records, information was recorded 

 relative to the elapsed time and tetiiperaturc and 



remarks were noted concerning the general com- 

 position of each samj)le. 



The percentage of dead embryos was computed 

 for each sample. Until stage 14 appeared this 

 percentage provided an estimate of mortality up 

 to the time of sampling since all embryos regard- 

 less of time of death were included in the computa- 

 tions. This procedure was possible since all eggs 

 used in any one experiment were recoverable 

 throughout the experiment. Although disinte- 

 gration of dead embryos did occur, the embryonic 

 membrane remained intact to the end of the ex- 

 periment. The information on dead embryos 

 per sample does not include data on time of death, 

 since stages could not be determined for embryos 

 that were decomposing, but it does provide good 

 information on the progression of mortality with 

 development. 



Because, as has been stated, the quantitative 

 sampling of lots of eggs become biased after the 

 appearance of stage 14 (hatching) and later stages, 

 none of the tables that carry details on individual 

 samples goes beyond the last sample that con- 

 tained stage 13 embryos (the bias starts, of course, 

 with the first sample containing stage 14 speci- 

 mens, but full records for stage 13 appear to be 

 desirable). In the experiments in which stage 14 

 was not reached, the tabulation ends with the last 

 sample that contained live embryos. Similarly, 

 the tables showing mortality by 1-day intervals 

 end with the records for the first day on which 

 stage 14 appeared, or, if that stage was not reached, 

 with the last day on which samples contained live 

 embryos. Terminal samples are included for those 

 experiments in which some embryos survived the 

 full term of the experiment or in which there was 

 cause to suspect that a few live embryos might still 

 be present. These terminal samples which con- 

 tained all eggs remaining at the end of the experi- 

 ment are considered, for practical purposes to be 

 unbiased. They were affected to some degree by 

 the earlier removal (after the start of stage 14) 

 of samples in which living embrj-os were taken out 

 of proportion to their true abundance, but the 

 numbers were so much greater in the terminal 

 sample than in these earlier biased samples as to 

 minimize the distorting effect. 



Records from biased samples were of course use- 

 ful for sliowing the time of first appearance and 

 duration of the later stages. 



