Ill SECONDARY AND MINOR INDUCTIONS 463 



(_/) Other in vitro inductions 



Grobstein (1955a) has chosen to study mammalian organ rudiments in his 

 investigation of induction processes. He has examined three instances of inter- 

 action between epithehal and mesenchyme cells of the mouse : chondrogenesis in 

 the vertebral column (see supra), the formation of the submaxillary gland, and 

 the development of the metanephros. The analysis is made possible, in the last 

 two cases, by the use of trypsin as a means of precise separation of two embryonic 

 tissues. 



The proliferation and especially the normal arrangement of the salivary epi- 

 thelium is the most exacting of these processes (Grobstein, 1953). It needs the pres- 

 ence of the capsular mesenchyme itself, to the exclusion of mesenchyme of other 

 origin (general mesenchyme, or mesenchyme of lung, or of metanephros). In the 

 case of the kidney (Grobstein, 1955b), the interaction is reciprocal; the mesen- 

 chymatous blastema of the mesonephric cord favors the proliferation of the ureteral 

 bud epithelium, while these cells cause the activation of the mesenchyme and its 

 organization into tubules. This last effect is not so specific. It is also obtainable 

 with the salivary parenchyme and the dorsal spinal cord'. It cannot be considered, 

 however, as a reaction indifferent to the nature of the inductor. Indeed, a rather 

 broad enquiry has been led with embryonic epithelia (salivary glands, lung, 

 stomach, prepituitary, pancreas, Wolffian duct), embryonic nervous tissues 

 (parts of the spinal cord, medulla, fore-brain, spinal ganglia), early embryo parts 

 (head fold ectoblast, neural plate, primitive streak, ectoplacental trophoblast), 

 and adult tissue (iris, cornea, retina, liver, kidney medulla, one kidney tubule). 

 All these living inductors, except the three cited as efficient, were negative. Simi- 

 larly, negative results were obtained with dead tissues, such as heat-killed or 

 frozen spinal cord, heat-killed liver and heat-killed kidney medulla. This is 

 strikingly different from the xeno-inductors which act on early stages (p. 427). 



Since spinal cord may be used for both vertebral chondrogenesis and metane- 

 phric tubule formation, a difference between the two cases must be stressed. In 

 the first case, the ventral part is more efficient, the dorsal one practically inactive; 

 in the second induction, it is just the reverse (Fig. 69, b). Thus, pointing to the 

 involvement of some ordinary cell constituent would be an oversimplification. 



It has often been thought that interposing some permeable membranes would solve the 

 dilemma between diffusion or contact influence. Grobstein has used this technique for the 

 submaxillary gland and for the metanephros, and got an ambiguous but interesting answer. 

 Using several "Millipore" membranes of 20 or 30 jj, thickness, with pores evaluated to be 

 0.5 (J.- 1 [J. diameter, he could show that the inductive stimulus passes through the membrane. 

 The salivary epithelium or the metanephric blastema is cultivated on the upper surface 

 of the membrane, the activating tissue on the under surface, and it is observed that the 

 reactor organizes much better, if not always typically, than it does by itself. Sections of the 

 membrane do not show any penetration, or only occasionally in the case of the salivary 

 mesenchyme (Fig. 96, a) while with the complex spinal cord-filter-kidney a penetration of 

 cytoplasmic threads is the rule (Fig. 96, b) ; if the pores are large enough, the threads can 

 be traced across the filter. By fixing in alcohol-formalin and scratching the tissue out, it 



' For the latter, the activity persists after dissociation and reaggregation (Auerbach & 

 Grobstein, 1958). 



Literature p. 483 



