384 JANSSEN ET AL. 



(Simpson et al., 1953), but this practice was discontinued when it was found 

 that adequate lesions could be produced through the intact skull. There- 

 after the operation consisted simply of incision and reflection of the scalp 

 prior to irradiation. The cyclotron aperture was circular and 14.3 mm in 

 diameter, sufficient in size to allow irradiation of most of the dorsal surface 

 of the cerebellum and cerebrum bilaterally. 



For morphologic studies, metallic impregnations for glia and Pickworth- 

 Lepehne benzidine staining for visualization of the vascular tree were used. 

 Usually the brains were fixed in Bouin's solution. Vascular permeability to 

 serum proteins was studied by the FLSP technique (Klatzo and Miquel, 

 1960). In this procedure, the rats were injected intravenously with 2 cc of 

 8% fluorescein isothiocyanate-labeled albumin, usually 24 hours prior to 

 sacrifice, and unstained, formalin-fixed, frozen sections were examined under 

 the fluorescence microscope. For observations on vascular permeability to 

 sodium fluorescein, animals irradiated at the same dose level were injected 

 with 1 cc of 10% sodium fluorescein intravenously, usually 24 hours prior 

 to sacrifice, and the gross observations for fluorescence were conducted as 

 previously described (Klatzo ct al., 1958). 



The alpha particles and protons to which the dorsal part of the brain was 

 exposed had an energy of approximately 12 Mev per nucleon. These par- 

 ticles were first deflected away from the magnet and down the deflector 

 channel, which contained a ^ mil Al stripping foil and a slit system to ex- 

 clude particles of unsuitable energy or charge-to-mass ratio. To obtain a 

 homogeneous particle flux distribution over the target area, a defocused 

 setting of a pair of cjuadropole magnets was used. The beam then passed 

 through a monitor, called a "high vacuum ionization chamber" (Brustad et 

 al., 1960) , and finally out of the vacuum system through a 1 mil Al window. A 

 specially constructed parallel plate ionization chamber (Brustad et al., 1960) 

 was attached to the snout, and the animals to be irradiated were placed in a 

 special holder about 5 mm from the end window of the ionization chamber. 

 This air gap reduced the effective particle range in tissue by less than 5 /x. 

 The ionization chamber could be detached and replaced by a magnetically 

 guarded Faraday cup. The monitor upstream was always calibrated against 

 the Faraday cup in terms of nimiber of bombarding particles per cm-. When 

 animals were exposed, the ionization chamber response and the calibrated 

 monitor response were recorded independently and converted to surface dose 

 in rad. These two difTerent dosimeters generally agreed to within a few per 

 cent. 



Between the ionization chamber and the monitor, sets of calibrated Al- 

 absorbers could be introduced with a remote controlled absorber changer. 

 With the Faraday cup connected, accmate determination of particle range 

 and beam-energy homogeneity could be performed routinely. With the 



