486 MAX MORSE 



wasser etwas NaHO zusetzt, namlich etwa 1,5 cchlttt NaHO zii • 



50 ccm. Seewasser" (Loeb '09 a, p. 158), which is distinctly not 

 possible with Cerebratulus in my experience. Nor in regard to 

 Chaetopterus, which Mead (I.e.) used at the very beginning of _ 

 the work upon artificial parthenogenesis, can it be said that 

 it resembles Cerebratulus in so far as reagents are concerned which 

 are efficient to induce development, for the addition of a small 

 amount of KCl is sufficient to start the spindle forming the con- 

 figurations of the anaphase and telophase of the first maturation 

 division.'* In the third case, Thalassema, it will be recalled that 

 Lefevre (I.e.) found that the following reagents induced develop- 

 ment, both maturation and segmentation: HNO3, HCl, H2SO4, 

 (COOH)., CH3.COOH and CO2; although the approach is much 

 nearer to Cerebratulus, yet there is a decided difference, since 

 Lefevre readily obtained trochophore larvae from 6 to 60 per cent. 

 Again, these reagents were efficient in themselves, while in Cere- 

 bratulus, no development was obtained without a combination 

 of reagents, except in a very few cases of saponin, used alone; but 

 here, no farther development than the 4 or 8-cell stage was 

 observed. 



Whatever the difference is — whether it be temperature, alka- 

 linity, the nature of the egg itself, being accustomed to lower tem- 

 peratures wherever it occurs than the other examples considered 

 above, so that the membrane may have become highly imperme- 

 able to ordinary reagents, and the egg nwdified to withstand un- 

 toward conditions — the fact remains that it is a difficult, and in 

 the light of the experiments described above, impossible task to 

 cause the development to proceed to larval stages. 



Experiments upon artificial hybridization 



The experiments which I have described in table 2 are some- 

 what misleading. Only one kind of each of the variations in 

 method of procedure is given and I have not made the complication 

 greater by the addition of any data as to the frequency of error. 



'' Godlewski ('11) however found that hypertonic solutions produced no effect 

 in this annelid. 



