LYMPHATICS IN FROG LARVAE 75 



spite of the great probability that the investigator will mis- 

 take primary lymph ducts lying close to the vein for out- 

 growths from its wall. This possible source of error was 

 pointed out in my report on injections in 1913-1914, and 

 again discussed by Kampmeier in 1925. It remains a serious 

 obstacle to the acceptance of Hoyer's theory. 



On the other hand, though these hearts are shown to arise 

 from vessels of the primary lymph plexus in both injections 

 and reconstructions, study of injections of early posterior 

 hearts in a series of frog larvae (figs. 33, 34, 35) has furnished 

 no evidence of such radical changes in relations between the 

 hearts and the plexus as are presented for the toad (Kamp- 

 meier). 



In the toad, the history of the anterior lymph heart is seen 

 repeated in the posterior hearts. Connections between the 

 lymph plexus and the heart gradually break off until the 

 hearts are isolated and reunion through new afferent ducts 

 is established later, just before the enlarged hearts join the 

 veins. 



The demonstration, by injection, of the development of the 

 posterior lymph hearts in frog larvae, on the contrary, fur- 

 nishes a straightforward history, without break in continuity 

 from their vessels or interruption of function, like the similar 

 history of the anterior hearts of these frogs. The history for 

 the toad is apparently exceptional. It is true that injections 

 of the posterior lymph hearts are not presented in as close 

 a series of stages as those in front, because posteriorly the 

 early hearts at first receive the injected fluid erratically from 

 the plexus in which they are developing. On the other hand, 

 the field of observation on the base of the tail is so favorable 

 that the operator sees many structural and functional features 

 which are not preserved in fixed specimens. Heart connec- 

 tions and heart function and influence on the plexus may be 

 thus checked from stage to stage, and the results selected in 

 figures 33, 34, 35, to represent three important stages, may 

 be accepted as a normal series (see plates 16 and 17). 



Figure 33 has already been described as typical of the first 

 stage of the posterior hearts. As development proceeds, the 



