— The appellant was convicted of manslaughter by culpable negligence in driving an automobile which collided with another driven by Samuel H. Roberts, on State Highway No. 11 about 13 miles west of Kirksville, on June 11, 1941, about 5:30 p. m. The punishment assessed by the jury was six months' imprisonment in the county jail. Both cars were badly damaged and Roberts died in a hospital in Kirksville that night about 2 a. m. There was no living eyewitness to the collision except the appellant, who stood on a demurrer to the State’s evidence. The evidence was all circumstantial. Appellant contends there was no substantial showing that he was culpably .negligent, and none that Roberts died as a result of the collision. It was the State’s theory that appellant was under the influence of intoxicants, and that he was driving at an excessive speed on the left side of the road. To give a correct understanding of the case we must state the facts in detail.
On the issue of intoxication. The appellant had been serving on a jury at Kirksville. Between 11 and 12 o’clock that morning he purchased and drank three glasses of 5% beer at the Cedar bar in Kirksville, and bought for $1 a bottle of apricot brandy which he. took with him. During the noon hour he drove deputy sheriff Shelby Allred about 7 miles west of Kirksville to see the Chariton River, which was out of banks. Witness Allred said that on the going trip appellant took one “swallow” of the brandy, and did the same thing once on the return trip. They were at the river about ten minutes and got back to Kirksville about 1 p. m. Nothing in appellant’s actions indicated he had been drinking — he didn’t act drunk. There was no evidence as to whether appellant ate a noon meal before or after the Chariton River trip. That afternoon about 3 o’clock he came to the home of Walker Simmons on Highway 11, about 17 miles west of Kirksville. There he met Russell Wilkins. They drove in *649appellant’s ear to the New Boston some 12 or 14 miles further south and purchased four bottles of beer of which each man drank one, the other two bottles being taken back to the Walker Simmons’ place, at which they arrived about 4 o’clock. Bernard Simmons joined them there and the three men drank the remaining two bottles of beer.
Wilkins did not see or hear anything of the bottle of brandy; Bernard Simmons was not interrogated about it; and neither was asked whether appellant was drunk or sober. But O. H. Kanan, a State trooper stationed at Kirksville, testified he called on appellant at the hospital shortly after 6 p. m., which would be a little more than 30 minutes after the collision. Appellant was in bed and had just come from the operating room where stitches had been made in a gash on his left arm and his eye had been bandaged. Appellant asked “Who is this?” Kanan told him, and appellant looked up and said “How are you, Kanan?” Then the witness inquired how he felt and was answered, “Not very good.” Next Kanan asked how much he had to drink, and appellant replied, “Too much.” He did not say when, but according to the foregoing testimony the last drinking the appellant did was about an hour and a half before the collision. The evidence does not show whether the beer purchased at New Boston was 5% or 3.2%, nor whether it was in quart or pint bottles; nor as to the quantity and strength of the brandy and what became of it, except two swallows. It also appears that during the period covered by the above narration, appellant drove 14 miles to the Chariton River and back; 16 miles from Kirksville to the Walker Simmpns place, and about 25 miles to New Boston and back.
Now as to the issue of excessive speed. After returning to the Walker Simmons place about 4 o’clock, appellant remained there until a short time before 5:30. He and Bernard Simmons were going fishing at the Chariton River, some 7 miles east on Highway 11. This required some preparation, such as getting fishing tackle' and dress-up clothes into which witness Simmons intended to change later. Simmons also maintained a repair garage on the place, and appellant had him examine a bent fender on his car, which he wanted Simmons to fix later. Simmons examined the underside of the car and noticed it and the whole car were clean and free from mud.
Appellant started out first and drove from the garage something like 100 yards over the private driveway to Highway 11 before witness Simmons had got his own ear turned around. When he reached the Highway appellant’s car was out of sight. But the road was hilly, and full of curves. Leo Wilkins was at the Walker Simmons place when appellant and Bernard Simmons left, and said appellant started about two minutes ahead. Capt. Ramsey from Jefferson City, of the State Highway Patrol, measured the distances along the Highway between places pointed out to him, the day *650before the trial, by the speedometer on his automobile. He testified it was 1.3 miles from the Simmons place to Allen Cook’s home, and 3 miles further to the point of collision, making 4.3 miles in all. But sheriff Love, who was with Capt. Ramsey on that occasion, and Lavaughn Cook, son of Allen Cook, and Bernard Simmons, said the whole distance was only 3 miles. Sheriff Love had lived in that country all his life, and the other two lived on the road. We take it that Capt. Ramsey was mistaken, and that the three other witnesses gave the correct distance.
.At any rate, the neighbor Allen Cook said he saw Simmons pass with another car immediately preceding; but he could not estimate their speeds. His son Lavaughn Cook gave what he called a guess that the car preceding Bernard Simmons’ car was running awfully fast — between 80 and 85 miles per hour, but at the end of the State’s case the court struck out this testimony. Bernard Simmons estimated it took him 4 or 5 minutes to traverse the distance. When he reached the scene of the collision the steam was pouring out of the radiators of the two automobiles. This is the only evidence as to excessive speed save the physical evidences, such as the violence of the collision, marks on the roadway, etc. Carl Noel, nearest neighbor of the deceased, who lived about 160 yards back from Highway 11 and about % mile east of the scene of the collision, saw a car pass at supper time, that he thought was the deceased’s. He estimated its speed at 35 miles per hour.
Next, we review the facts more nearly concerning the collision. The deceased was returning from Kirksville with stock feed and groceries. He was traveling generally west facing the afternoon sun, which set that day at 7:26 p. m., of which fact we take judicial notice. The road was straight and practically level for about 50 yards before reaching a moderate curve to the southwest where the road was slightly up-grade. The collision occurred a short distance after the deceased’s car had entered the curve. The direction of travel and the upward inclination of the road would make the sun shine through the driver’s side of the windshield. It was shinning that afternoon and would blind a westbound motorist more or less, according to the testimony of witnesses Bernard Simmons, Sherman Eitel and State Patrolman O. H. Kanan; though one State’s witness, Cecil Parmley, who was traveling west and arrived at the scene of the accident before the parties had been taken away, said the sun did not interfere with his driving.
From what has been said about the 50 yards of straight road before entering the curve from the east it must not be understood that distance was the limit of vision. The State’s photograph, Exhibit 1. taken at the scene of the collision, indicates that point could be seen for a considerable distance east of the straightaway. How far a motorist coming from that direction could see westwardly *651around the curve before he got to it is not shown. But the State’s photograph, Exhibit 2, taken from the scene of the collision and facing westward, shows the road extending for a substantial distance around and after the curve straightened out. Witness Bernard Simmons, driving east, said he could see the wrecked ears 150 yards away. There were some tall trees rather close to the inside (south) of the deepest point of the curve, which may have obstructed the view around it of a motorist traveling either direction; but no witness testified that was a fact.
Witness Simmons further testified there was a seepy, damp place at the south edge of the curving gravel road in the vicinity of the trees, for a width of 2 or 3 feet and a distance of about 50 yards, which caused eastbound motorists to drive more in the center of the road at that point. He said there was only one set of car tracks in the road, but other State’s witnesses said there were two. This seepy strip extended about ten yards west of where the wrecked cars were found. The graveled roadway was about 24 feet wide. On the north side the maintenance crew had left a “windrow” of loose supply gravel for repairs. Sheriff Love, who had been at the scene about an hour after the collision, said the north side of the gravel windrow then was about 4 feet south of the north edge of the graveled roadway — further south than as shown in the photographs taken by both sides. He said a road grader had swept the gravel north during the intervening time. The evidence does not show when the pictures were taken. O. H. Kanan said the gravel windrow was 5 or 6 feet out in the road the evening of the collision; and that the road is about 30 feet wide from shoulder to shoulder. Sherman Eitel was at the scene shortly after the collision, and made measurements the next afternoon at the request of appellant’s father. He thought the condition of the roadway was the same as the night before because he could still see the oil spots on the gravel road. He found the distance between the south shoulder of the highway and the gravel windrow was 20 feet. All these were State’s witnesses.
The gravel windrow, itself, was about 11/2 feet wide at the base, and according to sheriff Love’s version, which the State stressed at the trial, it shut off a strip about 51/2 feet wide on the north side of the highway, that being the right side for the. deceased as he traveled west. This left only 61/2 feet of the north half, or 181/2 feet of the whole road, open for travel. Deducting the 2 or 3 foot width of the seepy strip on the south side would reduce the width readily available for travel to 151/2 or 161/2 feet. So it resulted that the gravel windrow would have forced the south side of the deceased’s car very close to or over the center line of the road; and as the two cars passed they physically would occupy 12 feet of the 15% or 16% foot space of open road, leaving only 4% feet or less for clearance on both sides and between the cars. According to the other two witnesses that distance would be somewhat greater.
*652Bernard Simmons was the first person at the scene of the collision, Cecil Parmley was next. Sherman Eitel arrived soon after the two injured motorists had been taken to the hospital in Kirksville. State Patrolman Kanan and Sheriff Love came later, about 6 :30 p. m. At that time 4 or 5 cars were lined up at the side of the highway, and ears had been passing along the south side of the Highway between the two wrecked cars. The State contended the photographs offered in evidence both by appellant and the State, did not accurately represent conditions on the Highway immediately after the collision, because they do not show the two demolished cars in position, the wreckage scattered around, tire tracks obliterated by travel and the road grader, clay dirt in the roadway, and the gravel windrow in its proper place. Por that reason we do not set out these photographs but have inserted a sketch at the end of this opinion, which gives a general, but not entirely accurate, idea of what the State’s testimony showed.
The State’s witnesses all agreed that both the wrecked cars were standing upright. Appellant’s car was entirely off the graveled roadway on the south shoulder, headed southeast. ^ The deceased’s car was headed northwest, but more nearly crossways of the road, and about 10 or 12 feet north or northwest of appellant’s car. It was all on the north side of the center line (estimated, it was not marked) except the left rear wheel, which projected southeasterly 11/2 or 2 feet. Oil and grease had run out of the transmission and differential of deceased’s car, making oil spots in the gravel road which were easily discernible the next day and in the photographs. He was lying in the road 6 to 8 feet northeast of his car, with a bloody spot marking the place. The groceries and feed he had been carrying were scattered about the north half of the road around his car where it came to rest. There was an area about 5 by 10 feet on the north half of the road strewn with lumps of clay of rough shape except that they were flat on one side. The west end of this area was about 30 feet west'of where deceased’s car stood, and there were discernible tracks leading from it to the car.
Also, there was an oblong scooped-out place in the gravel roadway 3 or 4 feet west of this clay spotted area. "We so understand the testimony. In the examination of sheriff Love he was asked this question and gave this answer: “Q. Now, with reference to the scooped-out place, where was the clay dirt? A. Well, the scooped-out place must have been possibly' three or four feet ahead of the dirt — would be west.” Trooper Kanan testified it was west of deceased’s ear, but he did not say how far. It started 11/2 or 2 feet north of the center line and extended southeast 6 or 8 feet according to one witness and 21/2 to 3 feet according to another. Then it became shallower. There was another mark or dip in the gravel about 4 feet southwest of the scooped-out place. From these there *653were discernible tire marks leading southeast to appellant’s car. Trooper Kanan stepped the distance from the northwest end of the seooped-out place to the gravel windrow on the north and found it to be 3 steps, or 9 feet. From the same point to the south edge of the gravel roadway it was 4 steps, or 12 feet.
The State’s theory was as follows: Appellant had swung over to the north, or left, side of the road as he headed east around the curve. He collided with the deceased’s car while the latter was traveling west on his right side of the road. The crash came when deceased’s car was over the clay spotted area. It jarred the clay lumps from the running boards and bottom of the car to the road. The area was about the size of the horizontal outlines of the car. If this theory is tenable at all deceased’s car must have been headed southwest at the timé, as the damage to that car was at the right front corner and clear back along the right side, whereas appellant’s car was damaged mainly at the left front corner. We have not shown the clay spotted area as pointing southwest in the sketch because no witness said that it did, and all stated it was on the north half of the road. The scooped-out place in the middle of the road and the other mark near it were made by the front wheels of appellant’s car (the deeper one by the left wheel) as he swung it sharply back to the right at the instant of the collision. Appellant’s car ran on southeasterly, as the tracks show, to the south side of the road; and deceased’s car was pushed back or rebounded, as the tracks show, to the point where it stopped.
Ae few words more concerning the damage to the two cars and we shall have completed the statement of facts. The general location of the damage is shown by the shading on the two cars in the sketch. Appellant’s car was a 1935 Ford V8 coach, with only one door on each side. The bumper, radiator and headlamp, all on the left corner, were broken or damaged. The left fender was completely crumpled. The radiator was tilted back at the top and the engine block was broken at both the front and rear ends, together with appurtenant devices. The radius rods were broken loose from their attachment under the transmission. The left foot board was curled under for a foot or so below the left door, which, also was dented some, sagging and open. And the photographs show a dent in the left rear fender. But Bernard Simmons early in his testimony had stated the left rear fender was already dented when appellant was at his home between 4 and 5:30 that afternoon, before the collision; and that appellant had talked to him about rolling it out. So it seems the damage to the car did not extend back of the coach door on the left side. There was no direct substantial damage on the right side, and the glass in the windshield and left rear window was unbroken. But it was safety glass and not so fragile as ordinary glass.
The deceased’s car was a 1932 Chevrolet sedan with a lighter motor than the other. It was about demolished, especially on the right *654front and side. The front fender, radiator and engine assembly were pushed well over to the left. The axle was knocked loose. The right front door was dented and hanging on the top hinge with the door handle missing. The door post between the front and rear doors on that side was broken clear through about the middle. The rear door was torn off. The right rear fender was deeply dented in two places and the same was true of the right rear corner of the ear. The glass in the windshield and right rear window (which were not of safety glass) was shattered and mostly fallen out. The front seat was torn loose and projecting about a third out of the right side of the car. Both doors on the left side of the ear were closed and the handles intact. The hood cover was lying in the road southwest of the car, and the loose bumper in the same direction. The license plate was gone. One 100 pound sack of feed was lying in the road by the left hind wheel, and another broken sack was to the northeast. Groceries were scattered about the car.
The rule is settled that negligence to be culpable within the meaning of the manslaughter statute, Sec. 4382, R. S. 1939, Mo. R. S. A., sec. 4382, must be more than ordinary common law negligence. It must be so great as to indicate a reckless or utter disregard of human life.1 Neither can we convict appellant merely because he had been drinking; nor by an application of the doctrine of res ipsa loquitur.2 Further, we must be governed by the fundamental rules: that the appellant is presumed to be innocent; that the State must prove him guilty beyond a reasonable doubt; that he cannot be prejudiced by his failure to testify;3 that the difficulty of proving an essential fact does not dispense with the necessity of proving it; and that since the evidence is wholly circumstantial, the circumstances proven must be consistent with each other and tend to prove guilt, and must be inconsistent with any reasonable hypothesis of innocence.4
Measuring the evidence by these rules, we think there was substantial evidence that appellant was intoxicated — this because of his admission. Standing alone the other evidence was weak. The State proved only that during the five hour period of his imbibitions (preceding the collision about one and one-half hours) he had drunk at scattered intervals three glasses of 5% beer, two swallows of brandy and one and two-thirds bottles of beer. During the same period he had driven his car 55 miles. Deputy Sheriff Allred, who was with him during the noon hour when he drank the brandy, said he gave no evidence of intoxication. Russell Wilkins and Bernard Simmons, *655who were last with him, between 3 and 5:30 o’clock that afternoon, were not interrogated on the question. There was nothing in the first part of his conversation with trooper Kanan at the hospital indicating intoxication. But he further made the statement in answer to a question that he had drunk “too much.” This may give a clue to what he did with the bottle of brandy. At any rate all this evidence was competent in connection with his admission. State v. Campbell (Mo. Div. 2), 84 S. W. (2d) 618, 621(5); State v. Carter, supra, 342 Mo. l. c. 441(II), 116 S. W. (2d) l. c. 22(3).
The evidence of excessive speed was insubstantial, though admissible as indicating the speed at which appellant actually was traveling. With the testimony of Lavaughn Cook stricken out, the only evidence left was the testimony of Bernard Simmons and Carl Noel and the physical facts at the scene of the collision. Noel said appellant left the Walker Simmons place about two minutes ahead of Bernard Simmons. The latter stated it took him 4 or 5 minutes to drive to the scene of the collision. He evidently arrived there very soon after the crash. The radiators of the two cars were steaming. If the distance was 3 miles and the time 5 minutes, he averaged 36 miles per hour; if the distance was 4.3 miles (as Capt. Ramsey testified) and the time 4 minutes, he averaged 64.2 miles per hour. Considering the fact that the road was narrow, curved, hilly and had a gravel surface, together with the contradiction of Capt. Ramsey by the three local witnesses as to the 4.3 miles distance, the speed of 64.2 miles per hour can hardly be credited. But it indicates Bernard Simmons traveled at least 36 miles per hour, and he said he never came in sight of appellant on the way. The badly wrecked cars demonstrate that one or both had considerable momentum and therefore speed. Their respective weights are not given, but since appellant’s car continued to move forward after the collision whereas deceased’s ear was knocked back the inference is that appellant’s car had the greater momentum.
On the question of culpable negligence. The physical facts indicate appellant was driving on the north (left) side of the road. Indeed witness Simmons said the beaten track was more nearly in the center of the road at that point, and that he, himself, traveled there. The deceased also was driving on that side, though not clear over, probably because of the gravel windrow. It was his right side. There were no obstructions to prevent either motorist from seeing the other for a considerable distance. But neither turned out until too late to avert the collision. They waited until they were close together and danger was imminent, and then both turned toward the south about the same time, colliding cornerwise. Undoubtedly the left front corner of appellant’s car and the right front corner of the deceased’s car were the points of impact. The two ears showed that and tracks led from each car after it had stopped, back to the point of collision.
*656If two motorists with a clear view ahead and traveling at sub: stantiai speeds should continue driving toward each other on the same side of the road until they collide head-on, certainly no one would dispute that both were guilty of culpable negligence. We think it was only a little less so here, where they approached so closely that neither could avoid a crash by turning out. Under State v. Melton, 326 Mo. 962, 33 S. W. (2d) 894, this may have been true as to the deceased even though he was blinded by the sun. And he must have seen appellant finally; otherwise why did he turn sharply to his left? But he could not turn further to the right because of the gravel windrow if he was traveling at any speed. On the other hand the appellant had no right at all to expect a clear track and the sun was behind him. There was nothing to prevent his turning to the south (right) where he belonged. The very sharpness of the turns made by both indicate they were dangerously close. The facts warranted a verdict that appellant was guilty of more than common law negligence, and that his acts and omissions were in reckless disregard of human life.
These basic facts being established, oixr conclusion is unchanged notwithstanding we are unable to explain some of the other physical evidence. How did the deceased’s car roll back 30 feet, as the tracks disclose, when its front axle was knocked loose? What damaged deceased’s car clear back along the right side, even including two deep dents in the right hand rear corner, one of them high? There is no evidence of side-swiping, or that it turned over and rolled. With so great a crash why did the clay lumps on the bottom of deceased’s car fall perpendicularly to the road and stay there instead of being cast ahead? Why were the deceased and the heavy bags of feed not projected forward out of the car when it crashed, instead of being carried back until it came to rest and then thrown out?' However, on that point the evidence shows that the car stopped crosswise of the road with the open right side toward the place where the deceased and most of his cargo were thrown.
Nevertheless, we think the cause must be reversed and remanded for another reason. There was no direct proof whatever of the cause of deceased’s death. No evidence was introduced to show what injuries he sustained, or that they were fatal. In every criminal case the corpus delicti must be proven. In this case it was that deceased was killed through the criminal agency of another. 26 Am. Jur., sec. 461, p. 475; 30 C. J., sec. 360, p. 150; State v. Joy, 315 Mo. 1, 20, 285 S. W. 489, 495. The criminal agency was proven— the automobile collision caused by culpable negligence. But was the death produced by it? We do not mean to say such facts cannot be proven circumstantially, when necessary. State v. Smith, 329 Mo. 272, 278, 44 S. W. (2d) 45, 48(2). But the law demands the best proof which, in the nature of the case, is available. 26 Am. Jur., sec. 462, p. 476; 20 Am. Jur., sec. 1231, p. 1084; 30 C. J., sec. *657360, p. 150; State v. Henderson, 186 Mo. 473, 483, 85 S. W. 576, 578(2); State v. Poor, 286 Mo. 644, 659, 228 S. W. 810, 815(8).
The evidence shows deceased was lying in the. road after the violent collision; that there was a bloody spot there; that he was taken to the hospital; and that he died that night at 2 a. m. Appellant was in the same wreck; was injured and taken to the same hospital; and did not die. It would have been easy to produce evidence as- to the *658nature of deceased’s injuries, and to prove whether they were fatal — in other words, whether- he died of injuries received in the collision. The rule is ancient and for the protection of an accused. It must be followed.
For this reason the judgment is reversed and the cause remanded.
Leedy, P. J., concurs; Tipton, J., concurs in result.