Parts of the left tecta of the frogs were removed in one group. In another group, the left optic nerves of the frogs were cut through the mouth. The first group of animals survived for 12～26days and the second group 11～22 days. Both right and left retinas (with a small piece of optic nerve attached) were sectioned radially at 25?m on a freezing microtome and stained by Ebbesson-Heimer's method which allow degenerating fibers and terminals stained specially. We did not find any evidences showing degeneration. We concluded that centrifugal fibers coming from the centers to the retinas were absent in the frog.

The projection of the optic nerve to the visual centers of the frog was studied by means of autoradiographic and terminal degenerating method. The results obtained from the two methods were essentially identical in the distribution of the optic term inals in the lateral geniculate body and the pretectum neuropile contralateral to the [~3H]-leucine injected or enucleated eye. There are also some differences:1. In autoradiographic cross sections at certain level of diencephalon, the Bellonci nucleus contralateral to injected eye has a circular outline. It is filled with the optic terminals. Within the Bellonci nucleus and the lateral geniculate body ipsilateral to the injected eye, the distribution of the optic terminals is clear. The Bellonci nucleus contralateral to enucleated eye, revealed by degenerating technique, has a circular outline too, but the optic terminals mainly concentrate on the peripheral of the nucleus. In the central area of the nucleus there are only few degenerating optic terminals. And the Bellonci nucleus and the lateral geniculate body ipsilateral to enucleated eye have only few degenerating optic terminals. The outlines of the two nuclei can hardly be recognized.2. The optic terminals of the superficial layers of the tectum contralateral to injected eye appear to have random distribution revealed by autoradiographic method, while in the sections of the preparation with degenerating method, the optic terminals have a laminar distribution which is very similar to that obtained with modified Cajal's method.3. The appearance of the midbrain tegmental nucleus revealed by autoradiographic method was much clearer than that revealed by degenerating methodIn addition, it has been observed that the number and distribution of degenerating terminals varied with the animal's survival time after the operation.The causes which bring about the differences, the advantages and disadvantages of the two techniques were discussed.

In order to investigate the distribution and other features of callosal neurons inthe visual areas,12 adult albino rats were used in a series of experiments,in whichHRP(30～50%,DAB brown reaction)was injected into one hemisphere and theretrograde labelled cells were observed in the other hemisphere.1.Hardly any labelled neuron was found in contralateral visual cortices whenthe injections were made at sites within 3.6 mm from the midline of the brain.However,when the injections were made at sites 4～6 mm from the midline,manylabelled neurons could be observed.2.Generally,most labelled cells aggregated near the border between area 17 and18a.In area 18a numerous cells were labelled,which distributed in the form ofincontinuous patches.In the proper of area 17,however,less labelled cells werefound.3.A gross topographical correlation was observed between locations of injectionand labelling on the two hemispheres.4.The labelled cells were found in all cortical layers except layer Ⅰ.In area 17they were mainly in layers Ⅳ and Ⅴ,while in area 18 a they were found mainly inlayers Ⅲ,Ⅳ and Ⅴ.5.Some callosal neurons were lightly labelled.For these cells,only the outlinesof the cell body could be seen even under dark-field microscope.The diameters ofthese cells were about 8～10?m.A few cells were as large as 16?m in diameter.Other labelled neurons could be seen with both dark-field and bright-field illumina-tions.They were identified as pyramidal,multipolar bipolar and granular cells accordingto the shape of the cell bodies and the pattern of dendrites.The majority of labelledcells were pyramidal cells,with a long apical dendrite to the pia surface,10～20?min diameter.Other ceils were 10～20?m for multipolar cells,and less than 10?m forgranular cells.Comparisons were made between the above results and the physiological studies.

HRP solution (33%) was injected into the left superior colliculus of 21 adult albino rats. 1-2 days after injection, the animal was sacrificed. Brains were frozen sectioned and processed according to. DAB- or TMB-method. The labeled neurones were examined in the right superior colliculus and other areas of the brain. The resu ts are as follows:1. No matter where the HRP was injected (either in the rostral or caudal part of the superior colliculus or thepart between them) HRP labeled neurones were always observed on the opposite superior colliculus if the injection sites reached layers deeper than the stratum oPticum. The locations of the labeled neurones corresponded roughly to the sites of HRP injections. However, no HRP labeled neurones were observed when the core of HRP deposites was restricted to layers superficial to the stratum griseum intermediale.2. Of all labeled neurones, 53.6% were located in the stratum griseum intermediale, 16.5% in the stratum opticum. The rest were in deeper layers. In no case were HRP labeled neurones observed in the stratum zonale or stratum griseum superficiale.3. Labeled neurones could be classified morphologically into vertical fusiform, horizontal fusiform and multipolar neurones.4. In addition to the visual cortex, labeled neurones were also found in the inferior colliculi, paralemniscal nuclei, dorsal nuclei of the lateral lemniscus, substantia nigrae and lateral tegmental nuclei bilaterally. Labeled neurones were also found in the ipsilateral ventral nucleus of the lateral geniculate body, contralatera pretectal area and reticular formation.