Patterns of arterial supply to the atria were studied with respect to the coronary sinus region in 60 human hearts obtained from patients with a negative cardiac history and in 2 hearts with ECG-demonstrated coronary sinus rhythm. The atria, including the interatrial septum, were supplied by the right (40%) or left (30%) or both (30%) sinuatrial arteries which, however, did not reach as far as the posterior margin of the oval fossa and the coronary sinus. The region of the coronary sinus with the Eustachian ridge received direct branches from the trunk of the right (41 %) or left (11 %) coronary artery or both (48 %) in the adjacent part of the coronary sulcus. Possible development of the ‘coronary sinus rhythm’ on the basis of multifocal damage of the sinuatrial arteries is discussed.

Many authors, including Egorov (i 886), Beauvieux and Dupas (I 926), Winckler (I 93 I) Christensen (1936), Kurus (1956), and Grimes and von Sallmann (i960), have made anatomical and histological studies of the ciliary ganglion in man and various animals. They have given detailed descriptions of its macroscopical, topographical, and cytological properties, but without mentioning the blood supply. Only Kuznetsova (I963) has studied the vascular system of the ciliary ganglion in human foetuses and neonates. The present communication treats this subject in more detail, with special reference to an earlier study of the vascularization of the ciliary ganglion in the rhesus monkey (Eliskova, I 969). Material and methodsEighteen human orbits, four of them foetal, were studied. After the introduction of a cannula into the internal carotid artery, the arterial network was injected, first with a mixture of Indian ink and i o per cent. gelatin and then, after the dye had appeared in the conjunctival vessels, with another mixture consisting of barium and minium suspended in 7 per cent. gelatin. This was followed by an injection of IO per cent. formalin into the retrobulbar space. After the substance injected into the arteries had solidified, the orbit was exenterated and its contents fixed in io per cent. formalin. The vessels were traced under the binocular dissecting microscope, special attention being paid to the blood supply to and drainage from the ciliary ganglion. Four specimens of orbital contents were prepared for histological examination by serial sections in the sagittal and frontal planes and staining with Masson's trichome. ObservationsThe ciliary ganglion is a flat or lenticular organ with a medial and lateral surface, orientated in a sagittal plane, in close proximity to the ophthalmic artery (Fig. i) and lying just below its lower circumference and lateral to the optic nerve. Anteriorly, it may extend as far as the site where the artery bends to run across the nerve. Only in one case, in which the ophthalmic artery crossed the lower surface of the optic nerve, did the ganglion reach this angle. (However, this may be regarded as a variation in the course of the artery rather than as a variation in the position of the ganglion.) In six cases, the medial surface of the ganglion touched the posterior branches of the opthalmic artery, the central artery of the retina, and the posterior lateral ciliary artery.The ciliary ganglion is supplied with blood by one to four branches of neighbouring arteries. This was the posterior lateral ciliary artery (in eleven cases), the trunk of the

Lymph vessels were evaluated in 20 transplanted canine kidneys. Prior to the occurrence of morphological rejection changes, transplants with good blood flow rates show areas with dilated but, less frequently, also undilated lymph vessels. In transplants with the presence of a rejection infiltrate and a decrease of the blood flow rate of 21–50% the lymph vessels may be focally dilated or, in contrast, narrowed by the cells of the rejection infiltrate. Numerous vesicles and vacuoles can be seen within the cytoplasm of endothelial cells. Interendothelial junctions may be occasionally open, the walls of endothelial cells become attenuated, and disintegration of the cell membrane is followed by a focal destruction of the wall and the development of defects. Cells of the rejection infiltrate penetrate through the gaps in the wall of the lymph vessels. The possible mechanism of the disintegration is discussed in this paper.

The lymphatic and venous bed was evaluated in 42 human heart papillary muscles. The lymphatics form networks in the superficial and deep layers of the subendocardium and in the myocardium of the papillary muscles. In the apical area, the subendocardial lymphatics gradually merge with the fine myocardial, developing arcades at the origins of chordae tendineae. Thicker myocardial lymphatics leave the area passing along the connective tissue septa, often in close vicinity to the blood vessels. The injection method failed to reveal deep lymphatics in the connective tissue of chordae tendineae. Lymphovenous anastomoses were disclosed in the papillary muscle. The possible morphological pathways for lymphatic drainage of the papillary muscle are as follows: (1) the lymphatic bed subendocardial and myocardial, joining on the ventricular wall bed at the muscle base, and (2) lymphovenous anastomoses, passing along from the myocardial and subendocardial lymphatic bed into the ventricular wall veins and the sinus coronarius system, or into the thebesian veins and further directly into the heart ventricles.

Blood flow through the anterior interventricular vein was measured in 17 dogs under normal conditions and after ligating the descending artery. In 10 of these dogs retrograde blood flow was measured in the peripheral stump of the descending ramus and in the anterior interventricular vein 10 and 11 min later. After ligating the descending ramus 16.6–70.5% blood flowed through the ischaemic area into the anterior interventricular vein. The origin of the blood is two-fold: (a) arterial – blood reaching the ischaemic area via arterio-arterial anastomoses and the capillary bed into the vein; (b) venous – from veins of the nonischaemic area via venous anastomoses into veins of the ischaemic area. This flow was explained on a morphological basis by veno-venous anastomoses and by finding a venous pressure gradient between the ischaemic and nonischaemic area. In most cases (in 8 dogs), venous flow from the infarcted area was 14–91 % greater (average 47%) than retrograde flow in the peripheral stump of the artery. In two dogs only was venous blood flow the same or less than arterial. In casts of these two hearts, unlike the previous cases, the supply area of the descending artery and the drainage area of the anterior interventricular vein did not coincide and blood was partly drained into veins of the nonischaemic part. The results show that some of the blood flowing through the ischaemic focus is inadequate for the myocardium on account of its place of origin and consequent lack of oxygen. The possibility of influencing the retrograde arterial blood flow by ligating the septal artery or not is dealt with in the discussion.