Paper - The course of the blood flow through the fetal mammalian heart
|Embryology - 4 Aug 2021 Expand to Translate|
|Google Translate - select your language from the list shown below (this will open a new external page)|
العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt These external translations are automated and may not be accurate. (More? About Translations)
Kellogg HB. The course of the blood flow through the fetal mammalian heart. (1928) Amer. J Anat. 42(2): 443-
|Historic Disclaimer - information about historic embryology pages|
|Embryology History | Historic Embryology Papers)|
=The Course of the Blood Flow Through the Fetal Mammalian Heart
Howard B. Kellogg
Department of Anatomy, Northwestern University Medical School
Introduction And Historical
There is little exact information regarding the physiology of the fetal circulation in mammals. Two major problems present themselves: 1) the circulation of blood through the fetal heart; 2) the nature of placental exchange and the gaseous content of the maternal and fetal vessels. It is with the ﬁrst of these problems that the ‘present report is concerned, Whereas the second will appear in a separate publication.
The history of the dispute concerning the blood ﬂow within the fetal heart and its quantitative distribution dates from the beginning of the fourth century. It has been reviewed so thoroughly by Pohlman (’09 b) that only an outline and analysis of the more important phases are necessary.
1. Early discoveries
Galen (about 300 AD.) gave the ﬁrst adequate description of the fetal heart. He described the foramen ovale and the ductus arteriosus and mentioned their fates after birth. He further commented on the remarkable provisions of nature in requiring but a small amount of blood to ﬂow through the lungs before their respiratory function began. Wliile Galen’s account of the anatomy of the fetal heart was quite
1 Contribution no. 130. Submitted as one section of a thesis in partial fulﬁllment of the requirements of Northwestern University for the degree of Doctor of Philosophy.
accurate, he reversed the course of the blood and described it as ﬂowing from the aorta to the lungs by means of the ductus arteriosusi
At this time there was much mystery concerning the adult circulation, and Galen further erred in assuming that the blood somehow passed through, the ventricular septum on its course from the right ventricle to the left. Vesalius (1543), some time later, rightly questioned the teachings of Galen on this point.
Servetus (1553) made the important discovery that the blood, after passing through the lungs, is of a bright red color. It is possible that he might have interpreted this observation in terms of the pulmonary and systemic circulations had it not been for his untimely death at the stake as a penalty for his alleged radical teachings. Servetus also recognized the foramen ovalein the fetal heart.
2. The W0lﬁ”—Zt'egenspeck School
In the middle ‘ of the eighteenth century, VVolff (1778; Pohlman, ’09 b, pp. 79, 80) thought he had found a new anatomical relationship between theforamen ovale and the inferior Vena cava. He pictured the inferior Vena cava entering the two atria on their posterior walls so that the limbus of the foramen ovale cut the caval oriﬁce into two openings. On the basis of these relations, he taught that the inferior caval blood stream was split, half going to each atrium, and the effect being the same as if there were, indeed, two inferior venae cavae, one to each atrium. This conception means that the pulmonary return must equal the superior caval return, otherwise the circulation would be unbalanced.
Ziegenspe-ck.(’81, ’05) corroborated Wolff’s ﬁndings and likewise believed that the inferior caval stream was divided by the limbus of the foramen ovale. The following diagram represents Pohlman’s (’09 b, pp. 88, 89) quantitative interpretation of Ziegenspeck’s views: BLOOD FLOW THROUGH THE MAMMALIAN HEART 445
Lungs &———-——-————%—> Pulmonary return i
/' \. Right ventricle 1} Left ventricle 1} . \ /, Ductus arteriosus } Left atrium i \- /' Aorta descendens %—————-——>‘ Vena cava inferior §
Pars communis aortae & Right atrium %; /' \- ‘
Left ventricle 1} Right ventricle %
Aa. carotis et subclavius :} —-> Vena cava superior
These deductions were based upon a series of thirty—six injections into the‘ umbilical veins of dead pig fetuses and upon the analysis of a carefully measured calibration of the cardiac vessels of thirty—three fetuses. The injections indicated that one-half of the inferior caval stream goes to each atrium, and there are no objections to‘ such conclusions. It is Ziegenspeck’s interpretation of these results that seems to be at fault. V
Pohlman shows that Ziegenspeck’s Qonclusions are incorrect, for the reason that the laws of hydrodynamics do not permit vessels of the size listed to deliver the amounts of blood the arrangement calls for. Pohlman also makes the following criticism:
In any event the results will show that measurements of this character are valueless because We must grant that:
1. The lumina are exactly circular.
2. The vessel elasticity must be equal. _
3. The expansion of these vessels must be equal in all directions.
4. The intrinsic vessel resistance must be the same.
5. The capillary resistance in all vessels must be equal or known.
6. The quantity of blood expelled by the two sides of the heart must be the same and the pressure exacted equal. T
7. The vessels must undergo no particular change after death and
Hooker’s investigations (’21), in which he shows that the capillary resistance of a given capillary bed is constantly changing, makes it impossible to accept proposition no. 5, even though the area supplied by a given vessel might be 446 HOWARD B. KELLOGG
proportional to its carrying capacity. The marked difference in the histological structure of a vein and an artery makes it unsafe to assume that there is no change after death and ﬁxation, or, in case of a change, that it would have been proportional to the carrying capacity of the vessel. Pohlman further points out that Ziegenspeck entirely omitted the following:
1. The azygos circulation.
2. The coronary circulation.
3. The lymphatic return to the superior Vena cava.
3. The Sabatier School
Sabatier (1798) published his account of the blood flow through the fetal heart, and it is the prevalent one to—day. He describes the two caval streams crossing in the right atrium without mixing, the blood from the inferior Vena cava being directed through the foramen ovale by the eustachian valve to the left side of the heart and that from the superior vena cava passing through the tricuspid oriﬁce to the right ventricle. T
It is evident that such an arrangement would supply the head and upper extremities with blood richer in oxygen than that received by the rest of the body. The more rapid development of the head, upper extremities, and liver has been explained on the basis of a better blood supply to these parts.
Kilian (1826) believed that mixing of the caval streams occurs to some extent,_but he described an aorta cerebralis, which carried the best blood to the brain, and an aorta abdominalis, which carried the poorer blood to the rest of the body.
Reid (’35) performed three injection experiments upon dead human fetuses. The superior Vena cava was injected with a yellow mass and the inferior Vena cava with a red mass. No two of these experiments agreed. He accepted the results of one injection which did not show any mixing of the two ﬂuids, and remarked that, of the remaining two, one was injected in the wrong directionand the third was poorly BLOOD FLOW THROUGH THE MAMMALIAN HEART 447
handled. He thus satisﬁed himself that the head and upper extremities receive fresher blood than does the rest of the body. The objections to such work seem too great to require much comment. First, the substances used must have had a viscosity and speciﬁc gravity several times that of blood, or else they would have mixed to some extent before he could have opened the heart. Secondly, the heart, being dead, must have been more or less rigid, whereas the living heart undergoes dilatation while being ﬁlled. - He agrees that undoubtedly some organs do not require an especially rich blood supply for rapid development, but that, since the brain in the adult is very sensitive to minute changes in its oxygen
supply, so must it be in its growth. He further states:
And lastly when we consider the entrance of the two cavae themselves, the superior passing downwards and forwards, the inferior upwards and backwards, and to add to this the thick upper margin of the foramen ovale, we further perceive that the blood, passing down the cava superior must fall directly into the right auricle, to the left side of the Eustachian valve, and thus ﬁll the right ventricle.
To fulﬁll these conditions it is necessary to suppose that the fetus is kept in an upright position, or else the blood could not fall from the superior vena cava above to the right ventricle below. Pohlman ( ’07, p. 409), in his review of the Sabatier conception of the fetal circulation, raises these objections: .
A critical examination of the theory, which by the way is the prevalent one at this time, shows it to be physically impossible, morphologically inaccurate, and developmentally unnecessary. Physically impossible, because it would entail a distinct crossing of two currents under equal pressure in a distending chamber. Morphologically inaccurate, as Born has pointed out, in that no such. arrangement is found in the sauropsidian embryo, or even more strikingly in the human embryo, where the leg buds, once they appear, develop more rapidly than the arm buds when on the Sabatier principle the reverse should be the case. Developmentally unnecessary in that a number of organs grow rapidly from anlage to relatively large size without being vascularized (kidney, testes). 448 HOWARD B. KELLOGG
4. The Harvey-Pohlmcm School
The rather obscure theory of the fetal circulation given by Harvey (1628) seems, even to-day, to be correct in the main. Harvey formulated this conception after his epochal discovery of the adult pulmonary circulation andia careful study of the anatomy of the fetal heart. As Pohlman (’O9 b, p. 79) pointed out, the terminology used by Harvey has led to various interpretations. Pohlman’s careful translation of Harvey’s original description shows that Harvey believed the two caval streams to mix in the right atrium before passing to the ventricles.
For a time Mery (1645-1722) succeeded in overshadowing Harvey’s work with his idea that the fetal blood ﬂows from the left to the right atrium by means of the foramen ovale. However, Senac (1773), on repeating Mery’s injection experiments, proved them to be in error.
Pohlman. (’O9 b), utilizing pig fetuses, did the first work on living animals. He ﬁrst determined the capacities of the two ventricles and found them to be equal. This was done by slipping a ligature around the atrioventricular sulcus and drawing it tight at the end of the atrial diastole. The blood in each ventricle was then measured. Next, he demonstrated that the pressures in both ventricles were equal by introducing identical pipettes into the two ventricles at the same time. The blood mounted equally in both pipettes. He concluded that, since there was no oscillation of the blood at the time of diastole, there could be no suction or aspiration effect due to the expanding heart. The latter had been assumed to be true for some time, but had never been proved. ,
Pohlman then made a series of injection experiments, the technique of which will be found on page 451. He injected a suspension of cornstarch into the chosen vessel and withdrew equal amounts of blood from each ventricle at the same time. Seventeen injections were forced into the umbilical veins of pig fetuses. The blood samples drawn from the ventricles were arranged in pairs for the purpose -of comparison. In ﬁve pairs neither tube contained any starch. BLOOD FLOW THROUGH THE MAMMALIAN HEART 449
Since he opened the thorax after the injections were made, it is probable that in these cases the rapid fetal heart beat had already passed the starch through before it could be recovered. Twelve pairs showed an equal volume of starch in the blood from each ventricle, and this was conﬁrmed by rough microscopic examination.
Pohlman next succeeded in making seven injectionsinto the superior vena cava, although it was difficult to avoid serious hemorrhage. From these tests he obtained four pairs of tubes with equal amounts of starch inweach tube. Two pairs had a difference of over 50 per cent, «but there is no. record as to whether the results were consistent. One pair contained no starch. He further made simultaneous injections of colored starch into the superior vena cava and an uncolored starch into the inferior vena cava. After six such injections, the blood recovered from each ventricle was ﬁrst diluted in acetic acid to destroy the erythrocytes, and then compared. Both colored and uncolored granules Occurred in each sample. Quantitative counts could not be made for the iodine-potassium iodide used to stain the colored granules diffused too rapidly into the uncolored ones.
From these results Pohlman concluded that the caval streams do mix in the right atrium and that the foramen ovale does form a communication between the right and left atria, instead of being placed so as to split the blood from the inferior vena cava into two parts. He further declares that the blood to the head and upper extremities can be of no better quality than that to the rest of the body, and hence the more rapid development of the upper body does not result from a richer blood supply.
5. The present status of the problem
Inspection of many standard text—books of anatomy, embryology, and Obstetrics, since the time of Pohlman’s publication, shows a division of opinion between the Sabatier and Harvey—Pohlman theories. The former is favored by a large majority, several texts being non-committal in judg450 HOWARD B. KELLOGG
ment, and a few advocating the Harvey—Pohlman view. The Ziegenspeck-Wolff conception, based chieﬂy upon Ziegenspeck’s interpretation of his injections of dead pig fetuses and the calibration of the vessels of the heart, which since have been shown to be unreliable, is no longer considered seriously.
Regarding the two rival theories, it appears that the only experimental evidence in support of the Sabatier scheme is that of Reid, yet his work is practically valueless, because, of the three injections made upon dead human fetuses, no two agreed. The anatomical evidence supporting this view is based upon the position of the eustachian valve. This, however, does not seem to be adequate, for although in a ﬁxed, motionless heart the valve might be capable of directing all of the inferior caval blood through the foramen ovale, nevertheless, it must lose most of its positional advantage in the distending heart. The supposed morphological necessity of such a condition may be disregarded, since Born (’89) and Pohlman (’07,"p. 409) have argued convincingly from analogy that the more rapid development of the head and upper extremities does not necessarily depend upon a superior blood supply. On the contrary, it can be urged that the anatomical relations may be interpreted equally as evidence in support of either the Sabatier or Harvey-Pohlman theory.
The work of Pohlman, although highly suggestive and unique in its employment of experimentation on living animals, has lacked conviction, because his conclusions were based upon relatively few experiments. There is also a doubt as to their exactness, due to the theoretical disturbances of the circulation brought about by injecting a foreign substance into the heart and then piercing the heart to remove it. For these reasons and with the hope of gaining more precise information as to the degree of blood admixture in the heart, a program of experimentation was outlined. In this it was first decided to repeat, expand, and, if possible, improve on Pohlman’s type of experiments. It is with these results that the present paper is concerned. In a subsequent contribution more reﬁned quantitative determinations, using BLOOD FLOW THROUGH THE MAMMALIAN HEART 451
the methods of gas analysis which eliminate some of the more serious theoretical objections, will be reported.
I wish to express my great appreciation to Dr. L. B. Arey for suggesting this problem and also for his constant assistance during the work. The generous cooperation of Swift & Company, Chicago, expedited the work.
Starch and ink injections
The method of starch injections, as used by Pohlman, with some changes in technique was employed and a series of over 200 observations has been obtained.
Materials and methods. This type of experimentation is not dependent upon a normal placental exchange, and therefore permits the choice of pig fetuses as material. The large size and the availability of these fetuses is a distinct advantage in such work. In addition, one pregnant dog was secured and used while the placental circulation of both mother and fetuses was intact.
The pig experiments were done on the killing ﬂoor of Swift & Company, Chicago. Operative conditions were not the best, but the abundance of material compensated for any inconveniences. Fifteen to thirty minutes after killing, the sows were opened, the uterus cut at the cervix and handed to the experimenter. This lapse of time and the drainage of blood from the sow had a material effect upon the procural of suitable living material. About one uterus out of four of those large enough to use contained fetuses that maintained a sufficiently vigorous heart beat; the only blood samples retained were those taken from pigs whose hearts continued to beat at least ten times after the blood was withdrawn. When received, the uterus was opened at once along its least vascular border and the fetus removed with a minimum disturbance to the placenta and fetal circulation. It is possible to cut away the ventral thoracic wall and open the pericardium with but slight hemorrhage. In this manner the 452 HOWARD B. KELLOGG
heart is made freely accessible for rapid observation and
blood sampling before beginning the experimental procedure. It is important to eliminate the possibility of error through delay or by mechanical violence in exposing the ﬁeld of operation. ‘
The type of experimentation employed was the injecting of suspensions into the superior and inferior venae cavae and then either observing the visible effect on the two halves of the heart or making controlled computations of the contained blood-suspension mixture. The ﬁrst result is qualitative, but possesses the merit of furnishing ocular proof without the introduction of additional manipulative factors; the second is quantitative, and, properly controlled, affords a basis for deﬁnite mathematical comparisons. .
The technique of the direct-observation experiments was simple. Suspensions of India ink or of 10 per cent cornstarch in physiological salt solution were introduced into the selected vessel by means of a hypodermic syringe, whereupon the resultant color change in the heart was noted.
The objective of the quantitative tests was to obtain from many experiments a more accurate estimate of the extent of admixture of the two caval streams within the right atrium. This was to be done by introducing starch into a vein and then making counts of ‘the granules obtained from samples of blood withdrawn simultaneously from the two ventricles. Further checks were to be secured by comparison of the amounts of starch which settled out differentially from samples of blood allowed to stand in narrow tubes.
A 10 per cent starch suspension was introduced into the selected vessel by means of a hypodermic syringe and three to ten seconds later equal quantities of blood were withdrawn from the two ventricles with identical pipettes. It was necessary to use caution in the rate and amount of the injection to avoid excessive pressure within the vessel. The pipettes used were small pieces of glass tubing, held together by rubber bands, to which no. 20 needles were attached. The opposite ends of the pipettes were connected to a Y-tube BLOOD FLOW THROUGH THE MAMMALIAN HEART 453»
by means of rubber tubing. By sucking gently on the stem of the Y, a common negative pressure was imparted to each needle, and equal volumes of blood were usuallyobtained. Samples with appreciable differences in volume were rejected. The two samples thus obtained were then drained into paired vials, each containing ‘two drops of potassium oxalate. This amount of oxalate prevented clotting and was not sufficient to introduce any signiﬁcant error in the results.
When making a starch count, the blood was diluted one hundred times in a solution which contained a small amount of potassium iodide and iodine; the starch granules stained well with the iodine. Counts were made, using the chamber regularly employed in clinical blood computations. It was found that the specific gravity of starch granules is so high that they do not always spread evenly, but may tend to settle out on the slide near the point where the drop was emitted from the mixing pipette. To equalize this tendency, four counts were made from each sample, and the average value then used.
When blood containing starch is allowed to settle undisturbed, the starch granules, because of their greater speciﬁc gravity, occupy a bottom layer; next come the cellular ele -ments, whereas the highest stratum is of clear serum. This
differential sorting is preferable to separation by centrifuging. The volume of blood usually obtained in a single sample is 1 cc.,. but this amount is not sufficient to allow comparisons of the settled starch in ordinary centrifuge tubes. To overcome this difﬁculty, each tube received an equal quantity from eight to twelve corresponding pairs of samples. Composite columns of blood, 5 to 6 cm. in height, were thus obtained.
On the contrary, individual tubes were also prepared from
' glasstubing with a3-mm. bore. These tubes gave a column
of blood from a single sample ‘sufficiently high for the purpose of comparison. A large drop of mercury placed in the bottom of each tube produced a surface upon which the starch settled; this eliminated errors due to unequal curvatures at the sealed ends. 454 ' HOWARD B. KELLOGG
Observations. 1. Direct inspection of heart. Careful observation of the fetal pig heart showed that almost immediately after an injection of the starch suspension into the umbilical or external jugular vein both ventricles blanched. The apices of the two ventricles, being thinner than the rest, showed this decoloration very markedly. Similarly, when India ink was introduced into the veins, it was obvious that both ventricles became equally black at once. These injections were repeated on about 125 fetuses of various ages, and in every instance the discoloration occurred equally in both ventricles at the same moment. These observations, while not primarily of a quantitative value, constitute deﬁnite evidence proving beyond a doubt that the two caval streams do mingle in the right atrium, whereupon both ventricles receive mixed arteriovenous blood. They are especially important, because they involve no subsequent disturbance of the heart.
After the starch injections are made, the heart usually continues to beat at its regular rate, but frequently the heart is accelerated. This augmentation is probably due to the warmth of the starch solution or to a slight increase in pressure. Inthe case of the ink, which is toxic, the heart usually stopped within three to ten pulsations after the ﬂuid reached the ventricles.
Similar observations were made on ﬁve living dog fetuses, delivered under ether anaesthesia. The maternal and fetal circulations continued normally. The results from injections into the umbilical vein were identical with those obtained with pigs. The importance of such direct observations cannot be overemphasized.
2. Quantitative determinations. A. Results on pigs. a. Quantitative counts. Starch granules were counted in selected unit areas of the ruled chamber and the totals from ten fetuses summated. Such grand counts from pig fetuses that had received injections into their umbilical veins gave a ratio of 253 in blood withdrawn from the right ventricle to 293 in that from the left. On the basis of these results alone, the inference is that the left ventricle received 16 per cent more BLOOD FLOW THROUGH THE MAMMALIAN HEART 455
blood from the inferior vena cava than did the right (but see p. 458). _
12. Volume comparison. Three sets of composite tubes, each containing equal amounts from eight, nine, and twelve paired samples,’ respectively, were prepared from blood taken after the fetuses had received starch injections into the umbilical veins. From these a column of starch about 53 inch high Settled out. There was no appreciable difference in the amount of starch representing the right and left ventricles.
Thirty—three paired sets from umbilical—vein injections and thirty sets from external jugular-vein injections were next prepared. Each blood sample was placed in an individual narrow tube (compare p. 456). These occasionally showed a small difference in the amount of starch representing the right and left ventricles, but such variation was not consistently in one direction and the series as a whole essentially balanced. (Figures 1 to 3 illustrate well the type of result obtained from such injections.) The small bore of the indi~ vidual tubes made it diﬁicult always to introduce the complete sample into the tube. For this reason, the starch columns must in each instance be compared with the height of the corresponding column of blood. It is evident that in most paired samples the quantity of starch in each tube was proportionally the same.
B. Results on dogs. The dog fetuses were too small (although well toward the end of gestation) to attempt ex_ternal jugular-vein injections. Hence, the five fetuses were injected through the umbilical veins only and the samples placed in individual sets of small tubes. There was more variation in the corresponding tubes of this series (ﬁg. 3), the ratio between the right and left ventricle samples ranging from 10:11.7 to 5228.5, respectively; the average being 25.3): 33, respectively. Nevertheless, this difference is prob ably capable of explanation on the basis of technical procedure rather than representing a natural relation (p. 458). 456 HOWARD B. KELLOGG
Fig. 1 A series of ten pairs of tubes containing blood taken simultaneously from the right and left ventricles of pig fetuses after the injection of a starch suspension into the umbilical vein.
Fig. 2 A series of ten pairs of tube containing blood taken simultaneously from the right and left ventricles of pig fetuses after 3. starch suspension was injected into the external jugular vein. BLOOD FLOW THROUGH THE MAMMALIAN HEART 457
LR LR LR LR LR
Fig. 3 A series of ﬁve pairs of tubes containing blood taken simultaneously from the right and left ventricles of dog fetuses after the injection of a starch suspension into the umbilical vein.
EXPLANATION or FIGURES 1, 2, AND 3
As stated in the text, it was impossible always to ﬁll the 3-mm. tubes to the same level. Hence, for purposes of ready comparison, all columns of blood in the accompanying illustrations have been reduced to the same height, and the starch columns have been similarly corrected and reproduced in their exact ratios. . '
The letter L and R at the bottom of the tubes indicate that the blood was taken from the left or right ventricle. The solid block at the bottom of the tube represents the mercury put in to give an even surface for the starch to settle upon. The short white columns above the block repreent the sedimented starch, and the stippled strata correspond to the supernatant blood. 458 HOWARD 13. KELLOGG
The ocular observations of the starch and ink injections show clearly that the two caval streams do mix in the right atrium, while the starch counts and the sedimentation experiments are a quantitative proof that the two sides of the heart receive blood of approximately the same character.
The results, obtained in the pig series of over 100 injections and 125 direct observations of the heart, show some discrepancies, but they are not great and could, no doubt, be eliminated if a better control of operative conditions were possible (compare p. 451).
The heart beat of the fetal dogs was so rapid, due to their normal uninterrupted oxygen supply, that it seemed advisable to make the injection much faster than usual to prevent the starch from being pumped through the heart before the samples could be withdrawn. In doing this there is little doubt that the pressure in the inferior vena cava was raised considerably above normal. Since the foramen ovale is in a more direct line with the inferior caval blood stream than is the tricuspid oriﬁce, an increased pressure in the inferior cava would. tend to drive more than the usual amount of its blood through the foramen ovale. While these results showed the greatest variation of any in the present series, they nevertheless afford very important evidence in support of the Harvey-Pohlman theory, since they are the ﬁrst obtained on animals in which both the maternal and fetal sides of the placental circulation were intact and the other general conditions approached normality. M
The small percentage in favor of the left side of the heart receiving more pure blood, as indicated by the results obtained from the starch counts, may well be within the error introduced from the failure of the starch granules to spread evenly on the counting chamber. These results were all from the earlier injections, and again it is likely that the then undeveloped technique in controlling. the rate- of injection would account for more of the inferior caval stream going through the foramen ovale. BLOOD FLOW THROUGH THE MAMMALIAN HEART 459
Pohlman had difficulty in obtaining starch in all of the samples; seven out of twenty-nine sets contained no starch at all. This undoubtedly was due to the fact that he delayed too long his Withdrawal of the samples by opening the thorax after making the injections. By reversing these procedures in the present investigation, but three sets showing no starch were obtained in over a hundred samples. Pohlman also found great diﬂiculty in opening the thorax sufficiently high to expose the superior vena cava without severe hemorrhage ensuing; additional hemorrhage occurred as Well when the needle was Withdrawn after the injection had been made into the superior vena cava. By using the external jugular vein as an avenue to the superior cava, it was not necessary to open the thorax so high, and there was also no hemorrhage after the needle was "Withdrawn from, the small vein. The small size of the external jugular vein made only the larger pigs of value for this injection; many experiments Were rejected before the recorded numbers of successful injections were obtained.
The direct observations of the heart While starch and ink injections were made is so simple and reliable, for it eliminates the possibilities of induced error resulting from the trauma of pipettes, that it cannot be overemphasized as a critical method of attack. .
It is stated by some authorities that, as the fetus nears the end of the gestation period, the foramen ovale and ductus arteriosus become proportionally smaller and smaller with the concomitant increase in the blood ﬂowing through the lungs. If this were true, the fetal circulation would approach the adult type before birth. To gain information on this point, complete records were kept of the lengths of all pig fetuses. The pig fetuses varied from 6 to 14 inches in length. Pig fetuses 14 inches in length are nearly at term.
It was not possible to correlate any difference in the proportion of caval blood going to each ventricle with increasing size. A careful examination of a considerable number of fetal hearts seems to show that the foramen ovale maintains
THE AMERICAN JOURNAL OF ANATOMY, VOL. 42, NO. 2 460 HOWARD B. KELLOGG
its relative size throughout fetal life, but that the two atrial septa (primum and secundum) do become much larger in the late stages of gestation. Yet, while the septa do more nearly overlap, they are so placed that the blood may pass freely from the right to the left atrium until such time as there is an increased pressure in the left atrium to hold them in apposition. This, however, cannot take place until the respiration at birth forces a greater amount of blood through the lungs.
It will be remembered that Reid furnished the only experimental support to the popular Sabatier theory, but that his results are too meager and uncritical to be of any real value (p. 446); at best, the Sabatier conception is a mere philosophical argument, based on the supposed advantages that would accrue from a differential segregation of pure and impure blood.
On the contrary, the results presented in the present investigation agree exactly with those of Pohlman and attest to the fundamental correctness of the Harvey—Pohlman view. The fact that the present results were obtained from a large series of animals and several methods of attack (i.e., ocular observation, starch counts, and sedimentation of starch on living fetuses, some of which had normal placental relations) makes them much more trustworthy than those of but one method. If there is any difference in the quality of blood furnished to the different parts of the body, the amount is quantitative and not signiﬁcant.
To those who obtain mental satisfaction from ﬁtting supposedly advantageous mechanisms to physiological ideals, the Harvey—Pol1lman theory may be disappointing. Yet it is by no "means certain that there is any necessity in "maintaining a difference in the quality of blood distributed to the various parts of the body. Circulatory efficiency in the mammalian embryo doubtless depends jointly on the relatively large quantity of swiftly moving blood and on the large factor of safety in the oxygen supply. BLOOD FLOW THROUGH THE MAMMALIAN HEART 461
At the outset it was recognized that the injection of starch granules into the venae cavae as a method of attacking this problem has certain theoretical shortcomings. Objections to this type of experimentation apply especially to any quantitative deductions that may be made.
Possible errors might be expected to result from:
1. The introduction of a foreign substance into the blood stream.
2. Blockage by the starch granules of the capillary beds beyond the heart, which would result in stasis and churning of the blood within the heart.
3. The interference with heart action by piercing it when drawing samples.
4. Injection of starch suspensions under such great pressure that the equilibrium of the two caval streams entering the right atria would be disturbed.
5. Withdrawing samples from the two ventricles under unequal negative pressures, which likewise would change the natural circulation within the heart.
The objection to the introduction of a foreign substance into the blood stream can be minimized if the material used is not toxic and the viscosity and speciﬁc gravity are not at too great variance with that of normal blood. The starch suspensions in physiological saline are not toxic. It is true the viscosity of the solution is somewhat less than that of blood and the speciﬁc gravity of starch granules is greater than that of blood corpuscles. However, in a blood stream that is moving at a normal rate these diiferences are probably not large enough to have noticeable effect.
The possibility of vascular stasis resulting from blockage of the capillary bed beyond the heart'seems to be the most serious feature of this type of experimentation. Undoubtedly, stasis in the heart would result in churning of the blood and a consequent admixture through the foramen ovale of the blood in the two sides of the heart. That this did not occur
seems evident by the fact that no marked change in the heart.
rate nor in the amplitude of contraction and dilatation oc462 HOWARD B. KELLOGG
curred. It is well recognized that a sudden blockage of blood leaving the heart will cause dilatation, especially of the atria, a slowing of the rate, and contractions of a more forceful character. Further, the results seldom showed exactly the same amount of starch in the two sides of the heart and the Variations were not constant in any one direction. Had stasis and churning occurred, one might expect the samples from the two sides of the heart to contain equal quantities of starch. On the contrary, the inconstant Variation obtained seems to indicate a slight variable error due to different degrees of mixing in the heart itself.
Interference to heart action resulting from the introduction of sharp needles for the purpose of withdrawing samples can hardly be a serious objection, since the heart continued to beat regularly and normally and the samples were secured within two or three contractions after the needles were inserted.
Too rapid a rate of injection of the starch suspensions might introduce considerable error. As has been pointed out in the discussion of the results obtained with dog fetuses, too rapid injection into the umbilical Vein may result in more than the normal amount of inferior vena cava blood passing through the foramen ovale to the left side of the heart.
In any series of injections where such technique was employed throughout there undoubtedly would be a constant predominance of starch found in the left heart. This occurred in the dog series only. It might likewise be assumed that too rapid injection into the superior vena cava would result in a greater quantity of starch going to the right heart.
In the present series of experiments none of the superior Vena-cava injections was made rapidly, and since there was no predominance in favor of the right heart, it seems certain that these injections were not too rapid.
Since in the dog series only, in which injection was purposely much faster than in the pig series, was there a predominance in the amount of starch found in one side of the heart, it follows that the rate of injection, if properly controlled, is not necessarily a source of great error. BLOOD FLOW THROUGH THE MAMMALIAN HEART
Differences in the negative pressure applied to the two needles might cause the blood to be sucked by way of the foramen ovale from the side of the heart of least negative pressure to the side containing the needle of greatest negative pressure. This was overcome by the use of the Y tube connected to the two needles, as already described (p. 452). That the samples obtained simultaneously from the two sides of the heart were approximately always equal in amount is good and sufficient proof that this did not occur.
It is evident that the above objections apply chieﬂy to the quantitative results and, to a much less degree, to the ocular observations of the ink and starch injections.
Even though the theoretical objections to the injection method seem to have been sufficiently overcome in its practical application, fear that some might still be skeptical has led to a search for a new and more reﬁned quantitative means of approach.
The development of the Van Slyke micro gas analysis of blood for its oxygen and carbon-dioxide content has ideally satisﬁed this demand. This method makes it possible to determine with chemical exactness the amount of oxygen and carbon dioxide present in the blood of the various vessels of the fetus.
The results and technique of gas analysis will appear in a separate publication. It may be stated here that the results of gas analysis have shown beyond all doubt that the starchinjection method, when properly controlled, is a reliable and practical one. While it does not possess the mathematical exactness of the gas analysis, the conclusions drawn from its results are identical with those of the latter in regard to the course of the two caval streams through the heart. 464 HOWARD B. KELLOGG
After injecting India ink and starch suspensions into the external jugular and umbilical veins of living pig and dog fetuses,,direct observation and quantitative determinations prove that the blood of the superior and inferior venae cavae of fetal mammals is thoroughly mixed in the right atrium and then distributed to the two ventricles.
The foramen ovale is the same relative size in the later stages of gestation as in the early ones; the fetal circulation does not approach the adult type before birth.
The early rapid development of the head, liver, and upper extremities as compared to the growth of the rest of the body does not depend upon a superior oxygen supply.
Circulatory eﬁiciency in the mammalian embryo apparently rests on a relatively large amount of swiftly ﬂowing blood and upon a generous factor of safety in its oxygen.
BORN, G. 1899 Beitriige zur Entwicklungsgeschichte des Saugetierherzens.
Arch. f. mik. Anat., Bd. 33, S. 368, 369.
DALTON, J. C. 1884 Doctrines of the circulation. Lea’s Son & Co., Philadelphia. - HARVEY, W. 1628 Anatomical dissertation on the movement of the heart and blood in animals. Privately reproduced from the original edition printed at Frankfort-on-the-Main in 1628. With a translation and memoir for G. Moreton, 42 Bur-gate St., Canterbury, 1894.
HOOKER, D. R. 1921 Evidence of functional activities on the part of the capillaries and venules. Physiol. Rev., vol. 1, pp. 112-139.
KILIAN, H. F. 1826 ﬁber den Kreislauf des Blutes im Kinde, welches noch nicht geathmet hat. Karlsruhe, p. 200. (Loc. cit. Pohlman, 1909 b, p. 80.)
MERY 1645-1722 (Op. cit. Pohlman, 1909 b, p. 79).
POHLMAN, A. G. 1907 The circulation of the blood through the fetal heart. Johns Hopkins Hosp. Bull., vol. 18, pp. 409-412.
1909 The fetal circulation. Amer. Med., Burlington, Vt., vol. 4, pp. 317-320.
1909b The course of the blood through the heart of fetal mammale, with a note on the reptilian and amphibian circulation. Anat. Rec., vol. 3, pp. 75-109.
1916 The function of the foramen ovale. Interstate Med. Jour., vol. 23, pp. 105-124.
PREYER, W. 1883-1885 Specielle Physiologic des Embryo. Th. Griefen’s Verlag (L. Fernau), Leipzig, 8. 79-88, Tafel 5. r
BLOOD FLOW THROUGH THE MAMMALIAN HEART
REID, J. 1835 Injection of the vessels of the fetus to show some of the peculiarities of its circulation. Edinburgh Med. and Surg. Jour., vol. 43, pp. 11~13. 1835 Additional observations, Ibid., vol. 43, pp. 308—310. SA_BA'I‘IER., RAPHAEL B. 1791 Traité complet dhénatomie, T. 2, p. 493. (Loc. cit. Pohlman, 1909b, p. 80.) SENAC’ 1773 (Op. cit. Pohlman, 1909b, p. 79.) SERVETUS, M. 1553 Christianismi Bestitutio, p. 170. (Loc. cit. Dalton, p. 115.) VESALIUS, ANDREAS 1543 De Humani Corporis. Liber 4, Cap. 15. (Loc. cit. Dalton, p. 108.) WOLI"F,_ C. F. 1778 De foramine ovali ejusque in dirigendo sanguinis matro. Observ. novae. Nov. comment. Scient. Petropolit 20. (Loc. cit. Pohlman, 1909b, pp. 78-80.) ZIEGENSPECK, R. 1882 Welche Veriinderungen erfﬁhrt die foetale Hei-zth§tig— keit regelnfalssig durch die Geburt. Inaug. diss., Jena. 1884 P1-eyer’s Physiologie des Embryo, S. 596-607. 1902 ﬁber Foetal Kreislauf. Miinchen. 1905 Die Lehre Von der doppelten Einmiindung der unteren Hohlvene in die Vorhofe des Herzens und der Autoritéltsglaube. Sammlung klinischer Vortrige (Ser. 14, Heft 11), No. 401.