South Africa: Supreme Court of Appeal

You are here:  SAFLII >> Databases >> South Africa: Supreme Court of Appeal >> 1986 >> [1986] ZASCA 84

| Noteup | LawCite

S v Shabalala (450/85) [1986] ZASCA 84; [1986] 4 All SA 551 (AD) (28 August 1986)

Download original files

PDF format

RTF format

Bookmark/share this page

Bookmark and Share



In the matter between:

THE STATE Respondent


HEARD: 23 May 1986 DELIVERED: 28 August 1986


At about 2-30 a.m. on Sunday, 16 December 1984, an intruder, by removing a pane of glass from the bathroom


window frame and thereafter climbing through the resultant opening, gained access to the house of Mr and Mrs Allen. They, aged 34 and 29 respectively, lived in a residential suburb of Pietermaritzburg. Armed with a knife or similar instrument, he entered their bedroom. He approached the wife and, as she lay in bed, stabbed her in the left forearm, right shoulder and left breast. The husband, awakened by her screams, jumped out of bed and rushed at the figure that he saw. It retreated out of the bedroom. As Mr Allen pursued it, he too was stabbed several times; in particular in the left hand, the left forearm and, superficially, in the chest. The assailant fled back into the bathroom and escaped through the window by which he had entered. Mrs.



Allen was taken to hospital but died shortly after her ad-

mission there. On post mortem examination it was found that

her left lung had been punctured; the wound involved had

a penetration of 17cm. It, together with the loss of blood

resulting from the injury to the right shoulder, was the cause

of death. Mr Allen, consequent on treatment at the hospital.


It was these events that led to the trial

of appellant before KRIEK J and assessors, in the Natal

Provincial Division, on three charges, viz. (i) housebreaking

with intent to rob; (ii) murder and (iii) attempted murder

(of Mr Allen).

The State case was that appellant was the

intruder and perpetrator of the assaults. In seeking to prove

this, reliance was not placed on the evidence of Mr Allen,

who though an eye-witness to the occurrence, was not able

to identify the attacker. All he could say was that.

the person he saw had the build of a male of average

height, that he was dark-skinned and that he was wearing

dark clothing. Nor was there any evidence of fingerprints

having been found in or around the house. What the prose-

cution rested on was, in summary, the following. (i) The

finding by the police of what was said to be one of appel-

lant's canvas shoes (referred to as sandshoes) in, and the other

just outside, the house. (ii) The fact of his having, when

he appeared in a magistrate's court on Thursday, 20 Decem-

1984 in terms of sec. 119 of the Criminal Procedure Act,



51 of 1977, pleaded guilty, coupled with an incriminating

account by him of how he had committed the crimes. (iii) The

pointing out to the police, by appellant, of the Aliens'

house, together with certain places outside and in it. (iv)

The presence on the overall which he was found to be wearing

when arrested on Monday, 17 December (the day following the

occurrence) of (a) blood of the same groupings as that of de-

ceased and her husband, and (b) a hair similar to one from

deceased's head. (v) The giving of an alleged false ex-

planation as to his whereabouts on the night in question.

I deal in due course with appellant's evidence re-

levant to each of these matters. Suffice it at this stage

to say that, in support of his plea of not guilty before the

trial court, he denied that it was he who committed the

crimes. He testified that, though he had on the Saturday

(15 December) done some casual work at a house adjoining

theirs, he had never entered that of the Allens; he had

spent the night of 15-16 December in a hut situate on a

construction site in the vicinity.

The trial court, for reasons which will appear.

rejected appellant's alibi defence. He was, accordingly ,

found guilty but, seeing, so it was held, that neither an

intent to rob (nothing was stolen) nor to murder (Mr.

Allen) had been established, the convictions on counts 1

and 3 were respectively,of housebreaking with intent to

commit an offence unknown and assault with intent to do



grievous bodily harm. On each of these he was sentenced

to two years imprisonment. No extenuating circumstances

having, in relation to the conviction of murder, been

found, he was sentenced to death on count 2.

This is an appeal against such convictions and.

with leave of this court, also against the death sentence

It will be convenient to consider, separately, each of

the categories of evidence to which reference has been made.

I commence with that of the sandshoes ((i)

above). On his arrival at the scene at about 4 a m on 16

December, Lieutenant Upton of the Alexandra Police Station,

found one in the bathroom and the other lying next to the

outbuildings of the premises. It is plain that they had

been worn by the intruder (who had lost them, pro-

bably as he fled). Proof, therefore, that they were

appellant's would constitute damning evidence against him.

Though he did not dispute the State evidence that he had,

on the Saturday, been wearing shoes of a similar kind and

colour (which in fact fitted him) , he denied that they were his.

The State sought to establish the affirmative

by means of the following evidence. On Tuesday, 18

December, appellant was one of six persons who were

lined up in a row, one behind the other, in the yard

of the police station where he was being held in

custody. The one sandshoe, which had been retrieved

outside the house and which had not been handled by any-

one subsequent thereto, save that it had

been sealed in a plastic bag by the police, was pro-

duced and given by Detective Sergeant Collen to a dog

to sniff. It was no ordinary dog. It was a thorough-

bred English bloodhound, known by the name Tilly. This

type of breed possesses extraordinary powers of smell.

In the words of a Mr. Pead, a professional dog trainer

of 20 years experience (who was called by the State),

"they think with their nose"; they have "high level nose

power". They can, accordingly, be trained to track down

persons. This is possible because they are able to identify

a scent which is exuded from the body and becomes impreg-

nated in what is worn. Tilly had received such training

This took place at the police dog school in Pretoria.



At the end of an initial period of 6 months, she was

subjected to certain tests. They included tracking

and scent discrimination. The latter (which, for

present purposes is, I think, the important one) con-

sisted of the dog having to identify, from five aluminium

pipes, each of which had been handled, not more than 3

hours before, by a particular person (different in each

case), the one which "belonged" to the person whose hand-

kerchief or other item of clothing had been given to the

dog to sniff. This type of test was then repeated (though

whether with five other persons is not clear). Tilly's

identification was, on both occasions, correct so that

she obtained what is called an "A" certificate. Three



months later, having in the meantime been given

"praktiese werk" outside the school, she returned for

a second course at the end of which she underwent

trials of a similar kind save that this time what

was given her to sniff had been handled between 24

and 48 hours before. Again, she made no mistake in

her identifications of the pipes in question. She

then graduated with a "B" certificate. In the period

of one and a half to two years since then, she had been

used by the police to track and identify suspected

criminals. She had never been proved to be wrong.

On the contrary, in a number of cases where, there being

additional evidence, prosecutions had followed, convic-



tions had resulted. Collen had taken control of

her in the fifth month of the first session and had

been her dog master since then. Having sniffed the

sandshoe, Tilly, as she had been trained to do, walked

down the one side of the parade. Having reached the

last person, she proceeded round the back of him and

began to move forward on the other side towards the

front, at the same time sniffing each person she passed

When she reached appellant, standing in place no. 5,

she put her front paws on his shoulders and barked.

This indicated that his scent corresponded with that

which had been smelt in the sandshoe. The procedure

was twice repeated after appellant had chosen different



positions. The result, however, was the same. Each time the paw of suspicion was pointed at appellant.
In his evidence, appellant disputed that the parade had been fairly conducted. He alleged, in effect, that shortly before it, the dog had been brought into contact with him and that,during the identifications, she had been prompted. Collen's denial of these irregularities was accepted by the trial court. In my view, correctly so. The matter,accordingly, fell to be decided on the State version as set out above. It raised a problem which has engaged the attention of our courts on a number of previous occasions, viz., whether this type of evidence is admissible.



In R v Trupedo 1920 AD 58, following R v Kotcho 1918

EDL 91 and R v Adonis 1918 TPD 411, it was held it was

not. Here, too, evidence of the behaviour of a police

dog, which had tracked down the accused, after being

given the scent of certain footprints at the scene of the

crime, was in issue. Three reasons for its exclusion

are given. If the dog be regarded as the real witness.

hearsay evidence was involved. The dramatic nature of the

testimony might cause juries to attach a dangerously ex-

aggerated importance to it. But the main one, as I read

the judgment, is that its probative value being too tenuous,

it was not relevant. Thus INNES CJ, having observed (at

62) that "a fact is relevant when inferences can be



properly drawn from it as to the existence of a fact in issue" goes on to say (at 63/4):

"But to draw inferences from the actions of a trailing hound as to the identity of a particular individual is ... to enter a region of conjecture and uncertainty. We have no scientific or accurate knowledge as to the faculty by which dogs of certain breeds are said to be able to follow the scent of one human being, rejecting the scent of all others. ... The whole experiment ... contains too great an element of uncertainty to justify us in drawing inferences from it in the course of legal proceedings ; and evidence of the behaviour of the dog is therefore inadmissible."

Despite the (usual) clarity of the Chief Jus-

tice's words, the ambit of the ratio of Trupedo



has given rise to some controversy. A number of writers

have suggested that it does not lay down a general rule

that evidence of tracking by dogs is per se inadmissible;

the ruling had to be viewed in the context of the facts

of that particular case and especially the inadequacy of

general scientific knowledge on the subject in 1919; re-

levance, being a matter of degree, more convincing evidence,

including modern,technical information about the scenting

ability of dogs and their training, may justify its ad-

missibility, leaving only the weight of the evidence in

issue. (See May, South African Cases and Statutes on

Evidence, 4th ed., para 323; Schmidt, Bewysreg, 2nd

ed., p 356; Hiemstra, Suid-Afrikaanse Strafproses, 3rd

ed. ......,


ed., p.428; Barrie 1967 Codicillus, p. 44, and in particular L H Hoffmann: "Those Dogs Again" (1974) SALJ, 237) . Support for this approach is the persuasive authority of decisions by courts in a number of overseas countries in which this type of evidence is now admitted. The position ,in the United States is summed up in American Jurisprudence, 2nd ed., Vol" 29 sv "Evidence", para 378, as follows:

"There have been considerable uncertainty in the minds of the courts as to the reliability of dogs in identifying criminals and much conflict of opinion on the question of admissibility of their actions in evidence. A survey of the cases, however, reveals that most courts in which the question of the admissibility of evidence of trailing by bloodhounds has been presented take the position


that upon a proper foundation being laid by proof that the dogs were qualified to trail human beings, and that the circumstances surrounding the trailing were such as to make it probable that the person trailed was the guilty party, such evidence is admissible and may be permitted to go to the jury for what it is worth as one of the circumstances which may tend to connect the defendant with the crime."

(See, too, Wigmore on Evidence, 3rd ed, Vol 1 para 177). It has also been admitted in Scotland (Patterson v Nixon 1960 SCJ 42), Northern Ireland (R v Montgomery 1966 NI 120 which is criticised by F H Newark: "What the Dog Said", Vol 82 LQR 311 at 312) and by the New Zealand Appeal Court (R v Lindsay 1970 NZLR 1002), the British



Columbia Court of Appeal (R v Haas (1962) 35 DLR (2d)

172) and in England (R v Webb 1954 Criminal

LR 49).

The court a quo adopted this line. Moti-

vated, no doubt, by the fact that the evidence had not been objected to, the question posed by it was not whether it was admissible but "what weight, if any, can be attached to (it)". The conclusion arrived at was that:

"provided a proper foundation had been laid,
evidence that a dog which was given the
scent of some object chose one person from
among a group of persons as being associated
with that object, is part of the evidential
material which a court must have regard to
when considering the inference to be drawn
from the totality of the evidence led during
the trial (W)hether it makes a valuable



or an insignificant contribution to the totality of the evidence wil1 depend upon the circumstances of each case and the strength of the foundation which was laid."

A proper foundation was defined as including evidence as


"a) the handler's qualifications and experience;

b)the nature and duration of the training undergone by the dog;
c)the nature of tests undergone by the

dog before it 'qualified' and the results of such tests;

d)the dog's experience in doing this kind of work;
e)the dog's general skill and reliability, for example whether or not it had ever been proved to have been wrong;
f)the scenting ability of the breed to which the dog belongs;
g)the general basis for the suggestion that dogs generally, or dogs of a particular breed, have either an inborn scenting



ability or a scenting ability acquired by training;

h) the conditions under which the identification was made."

On the facts deposed to by Collen and Pead, it was held that such foundation had been laid for the reception of the evidence, which "justifies the conclusion that generally speaking (but not necessarily invariably) the scenting ability of an experienced and properly trained dog is reasonably reliable." The result was a finding that "Col-len's dog connected the Accused with the sandshoes", although, according to the judgment, it had no "decisive effect" on the verdict in the sense that it had not "tipped the scales one way or the other".

It would seem that KRIEK J, in his careful and



thorough judgment, did not go to the length of concluding

that the sandshoe was appellant's. Nevertheless, he did give the evidence of the behaviour of the dog towards appellant weight. The question is whether this constituted an unjustified departure from Trupedo, based on an unwarranted limitation of the ratio in that case. No doubt, the elevation of a particular decision on the relevance of evidence to a general rule has to be guarded against. Whether evidence is capable of inducing rational persuasion obviously depends on its probative force. And this can only be measured by a consideration of the facts of each case. Generally, only a pronouncement on law can constitute a ratio decidendi. (Hahlo and Kahn: The South African



Legal System and its Background, p. 260). "Decided cases

are ... of value not for the facts but for the principles

of law which they lay down", (per CENTLIVRES JA in R v

Wells 1949 (3) SA 83 (A) at 87-8). Nevertheless, when the

decision is that from certain facts certain legal conse-

quences follow, it is binding in any case raising substan-

tially similar facts. (Shepherd v Mossel Bay Liquor

Licensing Board 1954 (3) SA 852 (C,) at 861 A). Thus one

that evidence is relevant may lay down criteria that can

guide, or even be authoritative, in subsequent cases.

(LAWSA, sv "Evidence" vol 9, para 397 at p. 216). It

seems to me that, properly interpreted, Trupedo is an ex-

ample of the latter. This was, in effect, the view taken


by JACOBS J, as he then was, in S v Moya 1968 (1) PH

H148, and I agree with it. The judgment of INNES CJ

did not rest simply on a factual finding concerning the

reliability or otherwise of the particular dog whose ac-

tivities were in issue. In my view, it decided that, in

principle, evidence of the conduct of dogs, in identifying

an accused person by scenting, is inadmissible. The ap-

proval of R v Kotcho, supra, and particularly the mention

therein of the need for legislation if this sort of testi-

mony is to be admitted, makes this clear.

It does not follow that Trupedo is to be taken as

the final pronouncement on the matter in all circumstances

Despite the objection to the evidence based on its hearsay



nature, its exclusion is not absolute. It is still necessary to determine the parameters of the principle to be extracted from the decision. To do this, it is legitimate, and necessary, to look at the reason(s) underlying it (Pretoria City Council v Levinson 1949 (3) SA 305 (A) at 317). As already indicated, the principal one was the (extreme) untrustworthiness of the evidence. Where, therefore, this element is sufficiently reduced, even though it be not removed, the actions of the dog would become relevant and evidence thereof admissible. It is not possible to define what would have to be established to achieve this. However, it is apparent from the judgment that mere proof that the dog came from stock having special powers of discrimination



between the scent of one human being and another, that he was of pure blood and possessed these qualities himself and that he had been specially trained in tracking (being certain "safeguards" applied by those American courts which admit this type of evidence), will not suffice (see at 61-2). On the other hand, additional evidence explaining "the faculty by which (these) dogs ... are ... able to follow the scent of one human being, rejecting the scent of all others", would suffice. Whether the same applies to certain cases involving less convincing evidence is dealt with in what follows.

This being the broad effect of Trupedo, the next question, before returning to the facts, is whether it should, as was tentatively submitted on behalf of the. State, be departed . .

26 A

parted from. In my view not. The abolition of the jury system is not a good ground for so doing. The undue prejudice to the accused that was feared might result was but a subsidiary part of the reasoning. What was regarded as significant (at 61) was that no English authority favoured the admissibility of this sort of evidence. It would seem that, save for the isolated case of R v Webb supra, (a decision of the Hertfordshire Quarter Sessions, the brief report whereof gives no reasons for the reception of the evidence ) that is still the position. INNES CJ though, as indicated mindful of the approach of some of the American courts, was obviously not impressed with the safeguards referred to. And for good reason. They throw little light

26 B

on how the scenting process and alleged powers of discri-

mination work. Nor do they provide for proof that an indi-

vidual has, as far as dogs are concerned, a scent peculiar

to himself, a premise which, similar to the case of finger-

and footprints, is basic to the whole exercise. Naturally,

the distinction between admissibility and weight must not

be blurred. On the other hand, if the latter is so in-

consequential and the relevance accordingly so problema-

tical, there can be little point in receiving the evidence,

In my opinion, the cogency of the evidence in

casu, was not such as to remove it from the realm of con-

jecture and so qualify it for promotion to the status of

admissibility. In the first place, I have some doubt whether


26 C

the criteria, constituting the foundation as formulated by

KRIEK J, were satisfied. For example, though Tilly was

shown to be of good pedigree and well trained, no details

are, save to the extent indicated, given of her activities

during the period she was used after passing the second test,

There is a degree of vagueness in Pead's evidence as to how

long the scent on the sandshoe would have lasted. He con-

ceded that its strength diminishes with time. When two or

three days were suggested, he said "it should hold" - hardly

a convincing reply. There is also merit in the submission

of Mr Fuller, on behalf of appellant, that Collen's cre-

dentials, as a trainer, were not proved. As to the condi-

tions under which the identification was made, it will be re-

membered that the dog did not on each occasion sniff each



and every person on the parade; as I have said, she stopped when she came to appellant. Consequently the possibility that another might have had the same scent was not excluded.
Of more importance, however, is that, in any event the "proper foundation", though going further than the "safeguards", still falls short of what needs to be proved to render this sort of evidence admissible. Despite paragraph (g) of the requisites, it does not sufficiently provide for proof, and no evidence was adduced to show, that man's understanding of canine traits and capabilities, or, for that matter, their training, has advanced beyong what was known when Trupedo was decided. Nor can we make any assumptions in this regard, especially having regard to Pead's concession that "these being animals, we cannot understand the dog's nose ability". He did, indeed, express the opinion that a person's



scent (to a dog) is "as individual as fingerprints". His

reasoning, however, rested on a non sequitur which I need not

detail. He was, in the end, constrained to rely on the nega-

tive proposition that the contrary had not been proved. Perhaps

a sufficient inference that appellant had a scent different

from others could have been drawn, and the untrustworthiness

generally of the evidence reduced, by the use of more than

one dog at the parade (Pead admitted that this would have

made the identification more reliable) and by it, or them,

also being given the scent of an article which had been worn,

not by the suspect, but by another on the parade (to see

whether that person, rather than appellant, would then have

been "pointed out"). Another precaution might have been the

reholding of the parade from which appellant had been with-



drawn to test whether no one was identified (where the dog

had again been given the scent of the sandshoe). None of

these procedures were, however, followed. It was rightly

conceded that the fact that the dog had pointed out appellant

on three occasions did not enhance its probative value.

Pead, it is true, did say that the scent-discrimination power

of a correctly trained dog is "infallible". But this was

a bald opinion unsupported by acceptable reasons. Indeed,

the judgment a quo acknowledges that "there is in fact no

scientific basis for (this view)".

My conclusion on this part of the case is that

Trupedo was binding on the trial court, that it was not

distinguishable and that the evidence of the behaviour of



the dog towards appellant was inadmissible. It follows

that it should not have been taken into account and that the State failed to prove that the sandshoes were those of appellant. There was therefore no linking him to the crimes on this basis. Nevertheless, the fact that they were similar

to what he was wearing on the Saturday is of some significance in the general circumstantial picture concerning the identity of

the intruder.

I deal next with the probative effect, if any, of what appellant said in the sec 119 proceedings((ii) above). This section provides for the taking of a plea in a magistrate's court on a charge justiciable in the supreme court. The procedure involved is regulated by section 121 according to which, where a plea of guilty is tendered, the presiding magistrate must question the accused in terms of sec 112(1 )(b). If he is not satisfied that the accused



admits the allegations stated in the charge, a plea of not

guilty is entered and the matter dealt with under sec. 122

(1), "(p)rovided that an allegation, with reference to which

the magistrate is so satisfied and which has been recorded

as an admission, shall stand at the trial of the accused as

proof of such allegation". The reference to sec. 122 has

the effect of bringing into operation sec. 115 which, in

turn, enjoins the court to enquire from the accused whether

an allegation in the charger which is not placed in issue by

the plea of not guilty,may be recorded as an admission there-

of. if the accused so consents, such admission is deemed

to be one under sec. 220 with the result, in terms of the

latter section, that it is "sufficient proof of such fact"



The pith of what appellant stated, in amplification
of his plea of guilty, and in response to the magistrate
questioning him, was the following. Using a knife he
scraped out the putty holding one of the windows in its
frame; in this way he was able to remove the pane of glass
and then enter the premises through the opening, with the
object of stealing; he proceeded to the bedroom; as he en
tered , the two occupants awoke and rushed towards him; having
stabbed both of them, he fled. For reasons which it is un
necessary to canvass, the magistrate was not satisfied that
appellant admitted all the allegations contained in the
charges. Accordingly, and as he was obliged to do in terms
of the sections cited, he entered pleas of not guilty, where


after the proceedings were adjourned pending the decision of the attorney-general. Of more importance is that he first recorded, as a series of admissions, what appellant had said.
It was on these that the State, in the court a quo, relied. This it did by handing in the record of the sec 119 proceedings, as an exhibit under sec 122(4), read with sec 235(1) (which authorises proof thereof by mere production with the result that "any admission by the accused shall stand at (his) trial ... as proof of such an admission".) Clearly, if they did, they would have constituted evidence (or rather, probative material, as it is more properly termed - S v Mjoli and Another 1981(3) S A 1233 (A) at p 1247 fin - 1248 A) of decisive importance against appellant. No amplification of this proposition is required. Appellant,



however, contested the evidential value of his state-

ment (to use a singular, composite term for the plea

and admissions) . He did this on the ground that he was

coerced into making it. And in his evidence he testified

to certain assaults which he averred had been perpetrated

upon him whilst in police custody, in order to force

him to confess, and which, from fear of their re-

petition, had caused him to admit what he did to the magis-

trate. They (he said) took place on four separate oc-

casions and were committed by members of the police team inves-

tigating the crime. . The first was whilst he was

being driven in a car on Tuesday, 18 December by Detective

Sergant Njilo who, together with or in the company of two



other black policemen, twice during that morning administered

electric shocks to his body. The second (on the Wednesday

morning) occurred near a river to which he had been driven

by Njilo and the other two; again he was subjected to an elec-

tric current. On his return to the police station he was

taken to the office of Lieutenant Myburgh who then punched

him in his face. As a result the inside of his left upper

lip was cut. This was the third asssault. Finally, in the

afternoon of that day, having complained to a Mr. Leat, a

magistrate to whom he had been taken to make a written state-

ment, that he had been assaulted and having refused to do so,

he was, on his return to the police station, hit in his face

and kicked in the stomach by Warrant Officer Delport and Lieutenant


Upton. His statement the following day in terms of sec.

119 was not the truth. He made it because he "was tired

of being assaulted . . . There was nothing else I could do

because I knew that I will still go back ... The police

would kill me in the manner in which they had treated me.

I couldn't do otherwise". He did not tell the magistrate

that he acted under duress because the police involved were

present in court.

The State witnesses, viz., Njilo, Myburgh, Delport

and Upton, to whom these allegations of assault were put, de-

nied them. Nevertheless, the trial court declined to rely

on appellant's statement. The conclusion of KRIEK J was

that because of the "rather strange sequence of events ..


the safest course to adopt is not to attach any weight to what the Accused said during the Section 119 proceedings". The events referred to were not in dispute. In summary, they comprised the following "inconsistent" (as the learned trial judge put it) behaviour by appellant during the few days after his arrest: (i) On the day of his arrest, viz., Monday, 17 December he made an exculpatory statement to the police (to which I later refer as exhibit H); (ii) On the following day (Tuesday, 18 December),after the first alleged assault, on his return to the police station, he complained to a police Captain about his maltreatment and denied his guilt; (iii) On Wednesday, 19 December, however, after the alleged second and third assaults, he inculpated himself by


making the pointings-out referred to at the commencement of

this judgment; (iv) Later that day, as already indicated.

he refused to make a statement and at the same time com-

plained to magistrate Leat about having been assaulted; (v)

The following morning, after the alleged fourth assault;

he confessed his guilt in court.

Before us, Mr. Morrison, for the State, not sur-

prisingly, did not rest content with the trial court's

findings. His argument was that appellant should have been

held bound by his statement. The issue thus raised is not

one of admissibility. Having regard to the legislative

provisions referred to, there could have been, and was no,

objection to it being proved. At the same time, however.



it is clear, as the trial court recognised, that it did not absolutely bind appellant; he was entitled (as he did) to impugn its voluntariness and thus challenge its weight. This was so, irrespective of the exact status of the admissions forming part of the statement, i.e., whether they are properly to be regarded as formal ones under sec. 220, having the effect of dispensing with the need for evidence to prove the facts in question, or, though made coram curia in terms of sec. 121(2)(b), merely informal admissions forming part of the evidential material which became available to be used against appellant. (As to the difference between the two, see Schmidt, Bewysreg, 2nd ed., pp 214, 275). Dealing with the former, and in particular the meaning of "shall be



sufficient proof", RUMPFF CJ in S v Seleke en 'n Ander 1980 (3) S A 745 (A) at 754 F-H stated:

"Voldoende bewys is natuurlik nie af-ddende bewys (conclusive evidence) nie en kan later deur die beskuldigde, bv, weens dwang of dwaling of deur ander regtens aanneemlike feite, weerlê word."

The same applies, a fortiori, to less formal admissions (S v Sesetse en 'n Ander 1981(3) S A 353 (A) at 376 B-C). I shall assume (in view of a concession to this effect) that the onus was on the State to negative the alleged assaults; in other words, as in the case of the withdrawal of a plea of guilty at common law (in regard to which, see e g, S v Britz 1963(1) S A 394 (T) at 398 fin - 399 B), it suffices if the existence of the ground relied on to neutralise the weight



of the statement made in terms of sec 119 was, on all the evidence, a reasonable possibility. This would seem to be the position; see S v Tsankobeb 1981(4) S A 614 (A) at 624 H. There must,naturally, be a causal connection between the alleged duress and the making of the statement. This will not be assumed. As DE VILLIERS CJ said in R v Kumalo and Another 1930 AD 193 (at p 202), after having quoted Taylor on Evidence (10th ed, vol 1, para 866) that a plea of guilty in open court is "deliberately and solemnly made under the protecting caution and oversight of the judge" :

"We must assume that everything was properly done, that the admission was freely made, and that the accused fully understood and appreciated the consequences of his admission, just as we would



have had to make the same assumption, if the Court had accepted the plea and entered a plea of guilty".

With these principles in mind,I return to the facts. It would seem to be implicit that the true basis for not according appellant's sec. 119 statement any weight was not his inconsistent conduct per se, but a finding that the State had not excluded the reasonable possibility of appellant having been assaulted. Seeing that it is based on the trial court's assessment of a factual dispute, an appeal court, though entitled to, would usually not depart from a ruling of this kind. Here, however, it is justified. We have the benefit of certain credibility, or at least demeanour, findings of the court a quo. Upton,



Delport and Njilo were said to have "impressed us as com-

petent policemen who were giving an unbiased account of their

investigations. They made a favourable impression on us".

Myburgh was described as "giving an honest,factual ac-

count of (his) involvement in the matter with no attempt at

any embellishments." The judgment also records the

court's satisfaction "beyond any reasonable doubt that

(Myburgh's) evidence .... in all respects in which (he was)

contradicted by the Accused is to be preferred above that

of the Accused". On the other hand, not only was appel-

lant's demeanour held to be unimpressive, but,by reason of

(i) his general untruthfulness, (ii) his unsatisfactory

evidence in relation to the actual assaults, and (iii) the

probabilities ...


probabilities emerging from the evidence as a whole, it was concluded that "we do not have much faith in his account of the alleged assaults on him".
The record reveals ample justification for this evaluation. There are numerous illustrations of (i). As to (ii) and (iii), the following deserves mention. The injury which Mr Leat admittedly saw on appellant's face, viz. , a "slight break in skin on right part of the upper lip", i.e. on the outside thereof, is not consistent with appellant's description of the consequences of the third assault, namely, a cut on the inside of the left lip. Appellant conceded that the alleged fourth assault left no marks on his body. Having regard to its nature, it is



unlikely that it would not have. He said that the blow to the face felled him; he was kicked a number of times with such force that he had to plead with Upton to desist for fear of certain "stitches" on his stomach "bursting"; he was "writhing in agony". Unless the police knew that no' marks had been left, they risked exposure by taking him before the magistrate (and also having him medically examined by two different district surgeons viz., on the Wednesday, after the alleged third assault on him and on Thursday,after his appearance in court). It is a matter for comment, adverse to appellant, that he did not tell Mr. Leat of the second assault (which he described as the more painful). His excuse that he forgot is not acceptable. It is true that the State



evidence is not beyond criticism. The time that Njilo spent with appellant on the Tuesday morning does not seem
to be fully accounted for. The testimony of the two police-

men who admittedly accompanied Njilo on that day would have

been important on the issue of whether the assault took place. Yet they were not called as witnesses (though it must be added that it does not appear that they were available) . And, of course, as already stated, appellant made certain contemporaneous complaints about his treatment. Moreover,the medical evidence is that electric shocks would not necessarily have left marks on appellant's body. Nevertheless, on an overall view of the relevant evidence, I am of the opinion that the trial court should not have allowed



what amounted to a cloud of suspicion to obscure what con-stituted cogent reasons for rejecting appellant's allegations of assault and that in so doing, it adopted an over cauti-ous approach.
In any event, however, and whilst not overlooking that no reason - eg. , a realisation by him of the strength of the case against him - appears from the record as to why appellant should (voluntarily) have admitted his guilt, I am convinced that it should have been held that appellant's statement was not motivated by the fear of any physical violence to him. He did not, at least on the occasion of the final assault, testify to having been threatened with any consequences if he failed to incriminate himself.


48 Having, on his version, had the fortitude to withstand the

pressure until then, it is not explained why he felt com-

pelled to make the sec 119 statement. It is clear that appellant understood the nature of the proceedings and the consequences of what he said. The trial court regarded him as intelligent. In the circumstances, it is probable that, despite the presence of the police in court and the fact that he was not legally represented, he would have felt free, and indeed compelled, to seek the protection of the court - had he been acting under duress. His statement is a detailed one. His explanation that he fabricated most of it, or, to use his more colourful language, "I sucked it from my finger .... I was just talking", should have been rejected. Its similarity to the true events is too marked



for this to have been possible.

In the result, so it seems to me, the argument for

the State that account should have been taken of appellant's

sec 119 statement must be upheld.

This brings me to the pointings-out ((iii) above).

It was not in dispute that on Wednesday, 19 December, ap-

pellant directed the police to the Aliens' house where he showed

them inter alia the bathroom window (through which access to

it was gained) and the frames of certain other windows

(where the putty,holding the panes in place,had been tam-

pered with) and also certain rooms and spots inside

the house. This conduct, admissible in terms of sec. 218

(2), proved that he had knowledge of some fact relating to


what was pointed out (S v Magwaza 1985 (3) S A 29(A) at 39 G). On the facts of this case, this could only comprise the crime in question. The source of the knowledge might have been his participation in its commission or the fact that he saw others perpetrating it or information supplied by someone else (S v Gwevu and Another 1961 (4) S A 536 (ECD) at 537 E-F). Appellant's explanation fell into the last-mentioned category. He testified to having, the previous day, been taken to the house by Njilo (and the same two black policemen) and shown the various spots outside it which he later pointed to. From what he had been told by the police concerning the manner in which



the crime was committed, he was able to point out certain of the places inside it. In brief, then, he sought to make out the case that he had been precognized. Njilo denied the allegation. His version was that he had never taken appellant to the Allens house. The trial court resolved the resultant dispute in favour of the State, finding that the onus which rested on it of negativing appellant's evidence had been discharged. Accordingly, so it was held, it was to be inferred that appellant's knowledge stemmed from having been involved in the crime.
Mr. Fuller challenged this. It was submitted that if the trial courts could not and did not reject appellant's evidence that he had been assaulted by Njilo, it equally



should not have rejected his claim that he had been taken to

the house and "schooled". I am unable to agree. In my

view,there is no justification for disturbing what amounted

to a credibility finding. Though no specific reasons are

given for it, it must be examined against the background of

appellant's general untruthfulness and the favourable im-

pression that Njilo created on the court. The latter makes

it improbable that he would have indulged in what would have a-

mounted to grossly improper conduct. Furthermore, if with

knowledge of a contemplated pointing out, he was minded to

do this, he would surely have taken appellant inside the

house as well. Appellant admits this did not happen. One

of the spots pointed out in the house was a place in the pas-



sage, where, according to Mr. Allen, he pursued the criminal

as he retreated towards the bathroom. Appellant was unable

to explain what this pointing out represented. He was also

unable to satisfactorily explain how he was able to point

out (wrongly, as it turned out) the place where the pane of

glass was put after its removal from its frame. After

initially stating that Njilo had shown him where this was,

he contradicted himself by saying: " He pointed out the place

to me. He would not say where I had put it". This is, of

course, in itself improbable if Njilo was intent on precog-

nizing him. It must also be assumed that KRIEK J was alive

to what he would seem to have regarded as a question mark

over Njilo's evidence on the assault issue. In any event, he



should have, as indicated, decided it in Njilo's favour

The trial court, in convicting appellant, also

relied on the evidence of the blood and hair found on his

overall ((iv) above). I deal, firstly, with the latter.

The hair was found on the upper left front pocket of ap-

pellant's overall when it was examined at the police foren-

sic laboratories in Pretoria on 19 December 1984. It was

that of a white person and was similar in both colour (red)

and quality to the (head) hair of deceased. Indeed, accor-

ding to Major Oelofse, the expert who microscopically ana-

lysed it, the chances of it emanating from someone other

than deceased were 1:4500. Prima facie, therefore, it

gives rise to a fairly strong inference that appellant was

in .......

in deceased's presence, which, in turn, points irresistibly to him being the culprit (the hair having, in some way, got onto the overall as he stabbed her). There are, however, weaknesses in this reasoning. They arise from the facts being consistent with an innocent explanation for the presence of the hair on the overall. One is that there is merely the bald assertion that the statistic in question was based on research. Appellant's evidence that his overall was an external garment which he had worn often and for some time (whether in houses

where he was doing painting work, or when fre-



quenting public places, like shops), makes it reasonably
possible that the hair came onto the overall accidentally,
as he (probably) came into close physical contact with other
(white) persons there. A second is that the hair may
have got onto the overall via deceased's bedding (on which
her hair would be expected to be). This could have taken
place because there was evidence of two persons having, on
different occasions, handled both. Captain Van Dyk removed
the bedding from the bedroom and also received the overall
after it had been taken from appellant. Major Welma Oelofse,
of course, also dealt with the bedding and overall on their
receipt by her from the police. Thirdly, so it was sug
gested, the hair may have come onto the overall after Cap


tain Van Dyk had removed hairs from the deceased's body (which he did for their transmission to Pretoria for examination). I must say that I am inclined to think that the last two possibilities are somewhat remote and speculative; neither was put to the witnesses in question. However, it cannot be gainsaid that at least the first exists. On the other hand, I do not agree,as Mr. Fuller argued, that the evidence under consideration should have been disregarded. The trial court was entitled to put it into the scales against appellant. On the principle that it is not each proved fact which must exclude every reasonable inference save the guilt of the accused, but the facts as a whole (R v De Villiers 1944 AD 493 at 508), it had sufficient weight to warrant this.



The evidence of blood on the overall (which it

will be remembered appellant was wearing on his arrest and,

I may add, on Saturday 15 December as well) is even more

compelling. Spots of it were found near the left shoulder

and on the back. Its significance arises from the fact

that it belonged to groups A and NN, the same as those of

deceased (and Mr Allen) and different from appellant's.

According to the evidence only about 8% of the population

has blood of this classification. In the nature of

things, it is unlikely that blood (particularly some-

one else's) would come onto appellant's overall, at

least without him being aware of and remembering it.

The only explanation he could proffer was the unfounded and

far-fetched one that the police "planted" it there. This



suggestion was correctly rejected by the trial court. Again,the evidence under consideration does not, on its own, exclude every reasonable inference save that deceased's or her husband's blood (or both) came onto appellant's overall when he stabbed her (or him). Also, as was argued, the inference is not as strong as it would have been had deceased and Mr Allen each had different groups of blood both of which were found on the overall. Nevertheless, it is, to put it at its lowest, consistent with such inference (bearing in mind that both, and especially deceased, bled profusely). It was,therefore, correctly relied on by the court a quo as one of the pieces of circumstantial evidence implicating appellant.



Finally, there is appellant's evidence as to his

whereabouts on the night in question ((v) above). It was

rejected as false. This finding is unassailable. The

hut in which he testified he spent the night of 15-16 December,

was on what was referred to as the Haines construction site

But, the evidence of a number of lay, unbiased witnesses

as also that of the police, plus his own statement, Exhibit

H, properly construed, was overwhelmingly to the effect

that,on the day of his arrest,he showed the police a hut on

a different site where he allegedly slept, and not, as he

averred in evidence, the one on the Haines site. It was

common cause that he did not stay at the former. This re-

flects adversely on appellant's credibility. As was pointed


out, however, in S v Mtsweni 1985(1) SA 590 (A), caution

must be excercised in attaching too much weight to the fact

of an accused's evidence being untruthful. An innocent

person may falsely deny certain facts because he fears that

to admit them would be to imperil himself (S v Dladla 1980(1)

SA 526 (A) at .. 530 D). Nevertheless, it is a factor of

significance because appellant's evidence, in support of his

alibi, having been rejected, he is in the same position as

if he had given no evidence on the merits (R v Dhlomo 1961

(1) PH H54; R v Dladla and Others 1962(1) SA 307 (A) at 311 D-E)

That, then, is a survey of the evidence before

the trial court, its findings and our conclusions on the

individual correctness thereof. The result, in my view,


is the following. The finding that the dog connected ap-

pellant with the sandshoes was erroneous and constituted

an irregularity. It was not such, however, as to result

in a failure of justice. There was a formidable body

of evidence against him. In summary, it was his express

admission that he entered the house and stabbed deceased

and her husband (the sec 119 statement); the implied ad-

mission, flowing from his pointings-out, that he was in-

volved in the crime; the circumstantial evidence of the

blood and hair found on his overall; and his failure to

satisfactorily explain his whereabouts on the night in

question. There was also, as indicated, the fact that

on the Saturday he was wearing shoes of a similar kind, colour

and size to the ones found at the scene. One might, inci-

dentally, add the undisputed fact that appellant fell



within the general description of the intruder given by Mr

Allen, that he knew that the removal of putty from a window

frame would enable the pane to be taken out and that on 15

December, he worked, as a casual labourer, at a house ad-

joining deceased's; it apparently attracted his attention

because he admittedly noticed it. The only reasonable in-

ference to draw from the aforegoing, taken cumulatively, was

that appellant was the intruder and therefore the person who

committed the crimes of which he was found guilty. The

appeal against the convictions cannot succeed.

It remains to deal with the appeal against sen-

tence. There is no warrant for disturbing the finding that

extenuating circumstances were not proved by appellant.

It was argued that in entering the house, he did not anti-

cipate encountering anyone inside, that his possession of a



knife was merely to enable him to remove the putty, that his attack

on deceased (and Mr. Allen) was therefore not premeditated

and that when, to his surprise, he came across them, he

stabbed in panic; at that stage, deceased was (probably)

not asleep but had awoken and had confronted appellant;

there was no direct intent to kill but rather dolus even-

tualis. I agree with the trial court's rejection of the

argument. I am not sure that even if these facts were

proved, they would have sufficiently reduced appellant's

moral blameworthiness to constitute extenuating circumstances

In any event, however, they were not proved. There was, of course,

no evidence to this effect by appellant. Nor can they be

implied or inferred. There are no circumstances which could



have founded an assumption on his part that the house was

unoccupied. On the contrary, the impression to be gained

is that appellant was careful not to make an undue noise

as he entered; hence his removal rather than the breaking

of the pane. There is no reason to limit the contemplated

use of the knife to removing the putty. When he went into

the bedroom, he obviously then saw. the Aliens.Instead of has-

tily retreating, he attacked deceased. It is idle to

speculate what her exact position and state of conciousness

was when she was stabbed. I do not believe it matters.

It was never suggested by the State that appellant broke

into the house in order to kill her. Nevertheless, the

trial court, with justification, convicted appellant on



the basis of dolus directus.

The appeal is dismissed.