SCIENT. PAP. DESERT RES. STN, NC). 46 83

THE CALL OF THE BARKING ( : REPTILIA)

by W. D. HAACKE Transvaal Museum. Pretoria

(With 4 figures and 1 colour plate)

interpretations. Brain (1962), when revising the and describing P. carpi, repeated most of fie Gekkonjdae are the only with true them, adding his own observations and mentioning cal abilities and of this group the southern a call heard by Dr, Koch at Gobabeb, which had rican genus Ptenopux js probably one of the more clicks and followed in quicker succession than )st vociferous. The call mnsists of a series of is known for gnr~ulm.It was then assumed to haw quick ) clicks, ticks or chirps, which are uttered in been the call of carpi, but was later found by the 1 succession. 1 Calls are usually uttered in reply to author to be thaJ of koc?~i,a still undescribed 0 others, for which reason the type species was called in 2 1962. When describing kochi in 1964 the author

d garrulus. When describing this species in first mentioned a different call for each of the e t 1849, Sir Andrew Smith wrote: thrce species. At that stage it had aIready been a d noticed that not only a remarkable interspecific ( Tn the localities in which it occurs, many jndividuaIs

r may be seen peeping from their hiding places, each but also intraspecific variation in the calls existed e h uttering a sha~sound, somewhat like chick, chick; and that fluctuations in humidity and ternpr?rature s i and the number thus occupied is at times so great and l affected the reaction of She geckos. Since then an b the noise so disagreeable, as to cause the traveller to u change h is quarters." attempt has been made to obtain more detaiIed P

e information about the call of Ptenop~cs,especially h t At present three species are known, of which two P. gwrulus. The above-mentioned onomatopoeic re-

y were described within the last seven years. Each productions might, ta some extent. reproduce the b

d species has a characteristic call. Ptmopus is a sound and give some impression of the call, but, e t terrestrial genus and each specimen lives in a self- unless actual recordings are available, do not suppIy n a made burrow. The call is uttered by the male only any information regarding the duration, rhythm, r g

and the functional significance is still uncertain pjtch and other characteristics which are even more e c but probably is associated with territoriality and difficult to describe. Because of these shortcomings n e e attraction of females. Vocal activity usually an attempt was made to produce better comparative c i l arts towards sunset. reaching its peak just after means tape recordings subsequent material by of and r nset e before darkness falls. Silence usually foll- analyses. d vds, but loud choruses may continue right through n clicks u Each call consists of a series of uttered in

the night and isolated calls may be heard during y quick succession. Although Broughton (1963: 16) a the day. No calls are heard during the winter w and even in there may be silence defines '"the shortest unitary rhythm element of e months summer t fo. sound emission that can readily be distinguished as a r weeks. The ~rrelationbetween season, weather such by the unaided human ear" as a 'chirp', the G an ~dvocal activity is still not clear. t author wrefers the term 'click' for the individual e n sound is i Since the original description by Smith a number pulses of which a caII consists, as it a b have and better description of these staccato sounds. A con- a authors described the call Loveridge S

947: 35) summarized the various onomatopoeic siderabIe variation in the number of cLicks per call y b d e c u d o r p e R 84 DR. FITZSEMONS COMMEMORATIVE VOLUME

and the rate at which the clicks were uttered has of the Research Station at Gobabeb, while other been observed. observations were made during various trips to the northern Transvaal, Botswana, the northcrn Capc There has been some speculation in the past con- Province and South West Africa. In a number of cerning the sound-producing mechanism involved, places the frequency distribution of the number of and Loveridge (1947: 14 quoted by Brain 19621 clicks per call was established by writjng dawn the assumed that the sound was produced by "the number of cIicks for a long series of calls at a given sudden removal of the fleshy tongue from contact locality. This was done for P. garrulus only, as in with the palate'bwhile Goin and Goin (1962 : 267) the call of kochi the repetition rate of the clicks thought "that the sound is perhaps produced by is so high that it: is impossible to make accurate clicking the broad tongue agajnst the roof of the counts, while on the few occasions that the call mouth." As the larynx of gekkonids has membran- which is assumed to be that of carpi was heard, eous folds which can be considered to be vocal chords no recording material was avaiIable. (Keleman 1963 : 491 and Tembrock 1959: 37), this family is the only one amongst the lizards with true The recordings were a11 made on a NAGRA ID vocal abilities. The presence and structure of this portable tape-recorder and a Beyer M88 microphone organ has not yet been investigated in Ptenopm, was used. The recordings in the central Namib but, as this genus is so exceptionalIy vocal, it is Desert were made by placing the microphone on provisionally accepted that the larynx is well the ground about 12 inches away from the hole of developed and that the sounds are produced by It. the , while in the Gernsbok Park the micm- phone was used in connection with a 36 inch para- The function of the call is still uncertain. Mertens bolic reflector. The recordings were analysed on a (1946: 36) suggests that it is part of territorial Kay Electronic "Sona-Graph" 6061A while us in^ behaviour, but may also serve to attract females. a narrow band selector. A frequency spectrum of a This is possible, as no two holes were observed to signal Is thus obtained. Signals or parts of signals be closer together than about two feet, which may tasting not more than 2.4 seconds and a frequency indicate that females also show some degree of range betwecn 85 cycles and 8 kilocycles arc territoriality, which is unusual. However, as only r~cordedon a strip of spe-ial paper. This recording malcs have been observed calling, it is uncertain in is known as a "sonagram" (Andrieu 1963: 42) and what way females would define their territory. It by using the scale marked on The recording drum ) has been observed that the female of the Tokeh 1 of the "Sona-Graph'" the various time factors 1 (Gekko gecko) is attracted by the call of the male 0 involved can be measured. The total time for various 2 (Mertens op. dt.). During the middle of October calls and their "puhe repetition rate" (Broughton d e 1965 on the farm De Waal, Gobabis district, South The t 1963: 6) was thus established. first measur- a West Africa, just after dark, six specimens of ement was obtained by measuring the distance d

( were in P, garrulus colIected while running the between the beginning of the first and the end of r e road in the lights of an oncoming truck. At the the last cIick, while the second was estabIished by h s time other geckos were still calling. As all the i measuring the time-lapse between the beginning of l b spmimens collected were females, it is possjbIe that each pair of consecutive clicks in a call. For lack u this was a case where the females were responding P accurate of a more term one of the latter measur- e to the call of the males.

h ements is hereafter referred to as an "interval". No. t attempt was made at this stage to evaluate the y Hunsaker et al. (1968) found indications that b

certain geckos, for example Hemidactylecs, may not frequencies of each click. d e be able to hear their own calls, as their auditory t According to the Davis (1948) system the loci n thresholds are not in a direct reIationship to their a of the localities where these observations were made r vocaI ability. Although electro-physiological expe- g

are as follows: Annisfontein 1628 Ed, Daberas 1628 e riments have not yet been carried out with Pteno- c DC,De Waal 1823 Bd, Gobabeb 1523 Ca,Gorob Mine n pus, to check their hearing, behavioural observations e 1523 Cb, Hotsas 1522 Ca, Nossob Camp 2025 BC, c

i certainly showed clear reactions to the calls of other l Sendelingsdrif 1628 Bb, Swartpoort 1628 Eb.

r individuals. e d n u

DISCUSSION AND RESULTS

y METHODS AND MATERIAL a

w Ptenopus garmlus (Smith) e As Ptenopug rarely calls in captivity, all the t a recordings and behavioural observations had to be Under normal conditions garrulus calk in the G made in the field. Distress and agressive calls as late afternoon towards sunset, with greatest activity t e defined by Hunsaker et al. (1968) are ignored in just after sunset, quietening down rather abruptly n i

b this paper. Tape-recordings were only made at the as soon as darkness falls, However, especially during a Camp S Nossob in the Kalahari Gemsbok National rainy weather, loud concerts may continue through

y Park and in the central Namib Desert in the vicinity the night, only stopping after sunrise. Odd individ- b d e c u d o r p e R N

SCENT. PAP. W- nE3ERT EfW, NO. 46 ) 1 1 0 2 d e t a d ( r e h s i l b u P e h t y b d e t n a r g e c n e c i l r e d n u y a w e t a G t e n i b a S y b d e c u d o r p e R 86 DR. FITZSZMQSS COMMEMORATIVF: VOT-UME

EgnrrYluz N DIGorob Figure 1: Ijislograms showing variation in num- her ol cl~cks1wr call according to tape recordings. Top: Central h'amib Desert, Bottom: Nossnb Camp. Kalahari Gems- hnk National Park. 12 4 6 B 10 It 14 16 78 20

Gnboteb.5 of Uutseb raver

S )

1 0 2 L 6 8 13 '1 1- 16 18 20 1

0 Number of clrcks per call 2 d e t a d ( r e h s i l b u P e h t y b d e t n a r g e c n e c i l r e d n u y a w e t a G t e n i b a S

y Number of tlrchs Der call b d e c u d o r p e R SCIENT. PAP. NAMIB DESERT RES. STN, NO. 46 87

may be heard calling right through the day, but than at which they were recorded, the individual this is definitely a rarity. Only a single case of a clicks were counted and the resulting histograms day concert, as described by Smith (ap. cit.), has (Fig. l) show that most calk at Gobabeb consist been reported to the author. This took piace during of 5 cIicks, while most of those from near Gorob dull overcast weather (peT.3. cm, Dr. C. Koch). Mine and the Nossob Camp consist of 6 clicks. This No calls are heard during the winter months and r~sultgives the impression of a closer relationship even during summer the geckos may be quiet for between the last two populations although the first weeks on end (personal communications with ran- two are geographically much closer Srs of the Kalahari Gernsbok National Park}. Of the 12 calls recorded at Gobabeb 3 were Wrong may conclusions regarding their occurrence suitable for sonagram analysis, whiIe 6 of the 9 thus be reached, as for instance Mertens (1955: 51) calls from near Gomb Mine and 33 of the 239 did not hear calls at Swakopmund, Hentiesbaai and djstinguishable calls from the Nossob Camp were at the Waterberg, thereby assuming thejr absence. suitable for this purpose. The considerable intcr- In April 1963 the author heard qurr~lu~call at and intra-population differences existing in the Swakopmund, and specimens in the State Museum. rhythm only became apparent when studying thes~ Windhoek, pmve their presence at the Waterberg. sonagrams. The behaviour while calling has been observed The calIs from Gobabeb were very monotonous, and photographed on various occasions at localities i.e. all the clicks of one call have a similar pitch in Botswana, South West Africa and the northern and total amplitude and similar amplitudes at Transvaal and the same pattern was noted in all corresponding frequency levels (Fig. 2 b). The cases. Towards dusk the gecko emerges, takes up a rhythm js fairly regular and the average length position at the mouth of its burrow and surveys of the intervals for all the analysed calls is 0.39 the surrounding area with about half of the body seconds, from which the averages of the individual exposed. In this pose the yellow throat of the male calls differ from 0.01 to 0.05 seconds. The longest is well displayed (Plate I a-b). Usually in response total time for a 5 click call is 1.75 seconds. After to other calls, it will suddenly start inflating and a slow start the clicks follow in quicker succession, deflating its throat in quick succession, dip back but towards the end the intervals become more into the opening of the burrow until only the head extended again and the last is from 0.01 to 0.03 remains visible and calI rhythmically, opening and seconds longer than the previous one.

) closing the mouth for each individual click, 1

1 accompanied by little jerks of the head (Plate I b). The apulation near Gorob Mine has two distinct 0 calls, 0th of which consist of an average of 6 2

This action is reminfscent of a barking dog and as d sound is clicks (Fig. 1). The first type is regular and the e the produced certainly not a whistling one, t clicks are of equal pitch, total amplitude and a the name Barking Gecko appears to be more d duration (Fig. 2 c and Fig. 3 b), and appear similar ( appropriate than WhistIing Gecko ( FitzSimons

r to those recorded at Gobabeb. However, the average e 1943). h length of the interval is 0.27 seconds, which is 0.11 s In relation to the small size of this the i l seconds shorter than the average interval recorded b call is surprisingly loud and under ]deal conditions u can be heard for several hundred yards. for Gobabeb, and the total time of 1.45 seconds for P a call of 6 clicks from Comb is still much shorter e It appears that the volume is achieved by the h the for a t than 1.75 seconds call with only 5 clicks arnplif ying effect of the f unnel-shaped entrance of y from Gobabeb. The first interval of a call is also b

the burrow, into which the gecko withdraws to call. 0.01 seconds shorter than the consecutive ones, in d homing e It is also quite difficult to locate the caller by t contrast to those from Gobabeb where they are n jn on the source of the sound, as the voice appears seconds The a up to 0.02 longer. median intervals arc r to have ventriloquistic properties. By calling up- g equal in length and towards the end the tempo wards at an angle (Plate I b) the sound is scattered e domm of c slows again, resulting in an increase 0.01 over a wide area, which makes its place of origin n seconds in the last or each of the two Iast intervals. e

c difficult to locate. These acoustic effects were first i The second type of call differs considerably l from pointed out by Pmf. P. de V. Pienaar of the Depart- r the first. This is without a doubt the type of call e ment of Speech Therapy at the University of d described by Bradfield (Loveridge 1947: 35) as n Pretoria, when discussing sound anaIysis with him. u "squee - chi - chi", as the first click is particular-

y Tape-recordings were made south of the Kuiseb Iy emphasized. On the other hand, the first type a

w River at Gobabeb, about 3 miles NNE of Gorob Mine of call discussed above conforms to the monotonous e t in the central Namib Desert. and at the Nossob "gack - gack - gack" described by Brain (1962: a

G Camp in the Kalahari Gemsbok National Park. The 8) and accepted by Haacke (1964: 4). In the second t

e Popufations of the two first-mentioned localities type the initial click is louder (greater total ampli- n i belong to the g. maculatus while that tude) and more drawn out (about 0.05 secs. longer) b

a from the latter is of the typical g. gurmclus form. than the subsequent ones. There is a slight rise S While playing back the recordings at a slower speed followed by a sharp drop in the pitch at fairly y b d e c u d o r p e R P kodhi. Gobabab. ) 1 1 0 2 d e t a d ( r e h s i l b u P e h t y b d e t n a r g e c n e c i l r e d n u y a w e t rrne rn seconds. a , G t e

n Figure 2: Sonagrams of calls recorded in the central Namib Desert. i b a S y b d e c u d o r p e R SCIENT. PAP. NAMIB DESERT RES. STN, NO. 46 89

. 5 of riwn

eg,garrulus. Nossob Ca-- ) At sunset. 1 1 0 2 d e t a d ( r e h s i l b u P e h t y b

d eeqarrulus. Nnssob Camp. e t After midnight. n a r g e c n e c i l r e d n u y a w e t a T~mein seconds G t e

n Figure 3: Sonagrams of regular calls of P. gnrrulus. i b a S y b d e c u d o r p e R 30 DR. FITZSLNONS COLMME MORATIVE VOLUME

regular frequency levels over thz whoIe range of intervals from 10 analysed caIls of 4 geckos in this click. A similar picture, although to a lesser comparison with the 0.14 seconds of those recorded extent and over a shorter period, exists in the in the evening. Otherwise they are similar as far consecutive clicks (Fig. 2 d). On an average the as rewlar repetition rate and the delay of 0.03 to interval between click 1 and 2 is 0.27 seconds Iong. 0.04 seconds of the last click are concerned. AI- which is 0.08 seconds longer than the 0.19 seconds though the average interval for all calls fmm thp of the following intervals. The slowest time for a Gorob Mine area of 0.23 seconds is nearly identical caIl of the second type, with 6 clicks, is 1.23 seconds to that of the midnight calls this gives a wrong (average 1.11 secs., 3 calls), which is considerably impression. One can only compare the regular calls shorter than the 1.45 seconds for the calls of the of the first type with those of the midnight first type. In all the calls of the second type the recordings, but in those calls from Gorob lhrt last interval was again 0.01 seconds longer than average interval lasts 0.27 seconds with a resulting the previous ones. duration of 1.45 seconds for the total of a 6 click B3th typ~sof calls were heard at the same time call in contrast to the 0.12 seconds of those recorded of the day and in the same area. It was not ascert- at midnight. ained whether one specimen could produce both Of the calls recorded in the morning 7 were types of calls and an attempt b dig up a specimen analysed and the average for the 35 intervals was producing the second type was unsuccessful. 0.25 seconds, which is 0.03 seconds slower than thal During March 1967 at the new Nossob Camp recorded at midnight and 0.12 seconds dower than Mr. 0. P. M, Prozesky recorded 239 distinguishable that recorded during the previous evening. Although calls of P. g. gamlus, of which 141 were taken at odd calls during the evening and night had intervals and just after sunset, 63 between midnight and I which became progressively longer towards the end am. and 35 just before sunrise at about 6 am. of the call, usually about 0.01 second per interval, A light shower had fallen in the afternoon and the this is a fairly common occurrence amongst the geckos were particularly vociferous, continuing to morning caIls while the regular rhythm is the call right through the night, stopping only after exception and the sequence for one of the calls, sunrise the next morning. Mer a warm evening which is 0.26, 0.27, 0.28, 0.29. and 0.34 seconds, the temperature dropped noticeably and at midnight although exceptionally regular, can serve as a good as well as in the morning the caIIs were slow and example. )

1 drawn out and sounded as if the geckos were cold 1 The recordings made at the Nossob Camp give a

0 and tired (pers. corn. O.P.M. Prozesky). Although

2 good impression of the variation which can be number d varying in rhythm, the average of cIicks encountered at a given Iocality under varying e t per call remained six throughout the night (Fig. 1). conditions. The indications are that high humidity a

d Amongst the recordings made at sunset were those

( after rain will cause prolonged vocal activity, while

r of six specimens, of which from two to four calls a drop in temperature slows down the pulse repetit- e h each were suitable for analysis. The average intervaI ion rate. At present no exact correlation can be s i l between the beginning of consecutive clicks for made due to insufficient observations. The recordings b u these calls is 0.14 seconds, which is 0.09 seconds from near Gorob Mine show the variation of calls P shorter than that from near Gorob Mine and 0.16 which can be uttered by specimens of the same e h seconds shorter than that of calls from Gobabeb t population under identical conditions. A comparison

y and is very similar to the 0.13 seconds of the saPI of the recordings from the three different popula- b of P. kochi. The averages for the six individuals tions indicates to what extent various populations d e t vary from 0.13 to 0.15 seconds. In 9 calls the first differ. n interval was as Iong as the subsequent: ones, while a On two recent trips some additional observations r

g in the remaining 5 the first interval was from were made, which indicated that inter-population e 0.005 to 0.01 seconds shorter than the consecutive c variation is even greater than previously accepted. n ones. The median intervals of a call are of exactly e UnfortunateIy no tape-recorder was available so c i the same length, with only occasional minute l

that only the number of clicks per call for a num- r variations, while the final interval is usually 0.005 e ber of calk per locality was noted. In the Richters- d to 0.07 seconds longer than the preceding ones. In n veld of the north-western Cape Province calls were

u genera1, it can be said that the clicks in calls uttered analysed at the following four Iocalities: Annis- y during the evening are of similar pitch and foHew a fontein, Daberas, Sendelingsdrif and Swartpoort. As

w one another at a quick, reguIar rate with only a e the papulations in this relatively small area are all t slight delay between the two final clicks (Fig. 3 c). a considered to be of the subspecies P. g. maculatus. G

A remarkable difference exists between the calls the variation observed in their calls is quite con- t e uttered during the evenings and dose uttered at siderable {cf. Fig. 4). Similarly, on farm De WaaI n i The b midnight (Fig. 3 c and d). pulse repetition rate in the Gobabis district, South West Africa, calls a of the Iater recordings becomes much slower, with were heard which differed from those recorded at S

y an average of 0.23 seconds per interval for 51 the Nossob Camp (cf. Figs. P and 4). These two b d e c u d o r p e R SCIENT. PAP. NAMIR DESERT RES. STN, NO. 46

Figure 4: Histograms showing variation in numher of rlicks pcr call according to d~rectcounts. last-mentioned populations are of the typicaI sub- ation of the large pamphagid grasshoppers, such species P. g. gnrrulus. From the given histograms as Lamarcibhna. The number of clicks per call of it is obvious that the number of clicks per call 27 calls recorded at Gobabeb varied from 9 to 16, varies to such an extent that it is unsuitable to with 11 and 12 the most frequent numbers (Fig. 1). substantiate morphological subspecific differences. The average time-lapse between the beginning of consecutiv~clicks in 8 calls was 0.134 seconds. There Another impression gained on farm De Waal was is an indication that the pulse repetition rate in- that the most frequent number of clicks per call in creases in proportion to the number of clicks per a population on the one side of one of the long, call, as in a call. with 16 clicks the average interval consolidated dunes of Kalahari sand differed from lasts only 0.113 seconds, while in 3 calls with 11 that of the population occurring on the directly clicks each it lasts 0.148 seconds. TE this can be opposite side. As this gecko does not burrow in the

) proved to be constant, it means that the number of softer1 sand of the dunes and probably rarely crosses gecko. 1 clicks ,per call is predetermined by the this0 stretch of 1150 to 200 yards, these dunes may of this 2 Although the calls species have a great be important barrjers. As they may be several miles d number of clicks, an analysed caII of 16 clicks lasted e tongt they could even be as effective as river-beds, only 1.775 seconds (Fig. 2 a), which is not much a asd was noticed in the central Narnib at Gobabeb. longer than the 1.75 seconds of a 5 click call of ( only one Unfortunatelyr evening could be spent at gu~arruEw~recorded at the same tjme and at the same e

Deh Waal and the difference in the calls on the other locality. The pulse repetition rate is so high that it s i sidel of the dun^ was only noticed very late. after is practica1Iy impossible to recognise and count the theb main period of vocal activity. u individual clicks of a call in the field. This caused P the erroneous statement by the author (1964: 4) e Thjs observation suggests that the number of h Only t that the caIl consists of 9 to 11 clicks. when clicks is likely to be geneticalIy fixed in a population. y studying the sonagrams and replaying the record-

Whereb the distribution of this gecko is continuous ings at a slow speed was it reaIised that the actuaI gened flow can occur, resulting in graduaI changes e int the calls, while barriers, such as mountain numbers were usually higher than previously accept- n

a ed. The pulse repetition rate usually starts off slowly,

ranges,r dunes and river-beds, will cause a discon- g i.e. with a long interval between click 1 and 2, then tinuity resulting in noticeably djfferent calls. On e thec other hand it is possible that the number of picks up speed, j.e. the intervals become progressive- n e 1y shorter, and slows down towards the cnd. The

cljcksc per call in a population Is imitated by and i l following series of intervals from a 13 cIjck call imprinted in juveniles, which would also result in r ae certain degree of homogeneity within a given jllustrates this clearly: 0.160, 0.155, 0.150, 0.150, d 0.145, 0.140, 0.140, 0.140, 0.135, 0.135, area.n This couId be verified by transferring young 0.140, 0.140, u specimens to other areas where they could grow 0.140 seconds. y upa under observation in close contact with a diff-

w The behaviour while calling is similar to that erente call. t descrjbed for garruTu~and its position in relation a to its burrow, before and while caIIing, can be seen G

Ptenopzcst koclti Haacke in Plate I c and d. The call is also loud and. under e n i The call consists of a series of clicks of apparently favourable conditions, can be heard for a few hun- b dred yards. It is difficult to pinpoint the exact jdentjcala pitch and loudness, uttered in extremely S quick succession and sounding similar to the stridul- position of a gecko by listening to its call because y b d e c u d o r p e R 92 DR. FITZSLWDNS COMMEMORATIVE: VOLUME

of the habit of calling upwards from inside a small varies according to season, time of day and crater, thereby scattering the sound. OnIy males prevailing weather conditions. have been observed calling. Although vocal activity 7) The pulse repetition rate and number of clicks usually commences just before sunset and reaches per call of 7cochf is much higher than that of its greatest volume just after sunset, calling may garrulus in the same area. continue right through the night and on rare occa- sions, especiaIIy on overcast days, individual calls may be heard during the day. According to Dr. C. h Koch and Mr. J. du Preez (pers. corn.) no calls 2~ are heard during the winter months at Gobabeb. E S c5 \ The X call of this species was first noticed at Go- AV.length babeb S .z by Dr. C. Koch, whose observations were 8) Locality 4 intervals Duration of calls published by Brain (1962: 15). It was then ten- tatively attributed to P. m*. At a later stage the P. g, garrulsa author found that this was the call of P. kochi, a species stiIl undescribed in 1962. Nossob Gamp Evening 6 0.137 sec. 6 clicks 0.74 sec. Ptenopw carp%Brain Midnight 6 0.225 sec. 6 clicks 1.26 sec. Morning 6 0.254 sec. 6 clicks 1.51 sec. No specimen af this species has yet been observed De Waal 4 - while calling. In October, 1963, on the gravel plains north of the Kulseb River, calls were heard which P. g. m.a.culatus were quite different from those of garrulm and Gobabeb kochi. They consisted of monotanous series of up S. of river 4 0.390 sec. 5 clicks 1.75 sec. to 16 clicks, which were uttered at a very slow but Gorob Mine constant repetition rate and with a rather low pitch. TYPe a) 6 0.270 sec. 6 clicks 1.45 sec, It was already too dark to see geckos calling. A few TYFje b) 6 0.200 sec. 6 clicks 1.23 sec. moments later, with the aid of a flashlight, a spe- Annisf ontein 4 cimen of P. carpi was found in the area from where Daberas 3 these calls were heard, and the caIIs were attributed ) Sendelingsdrif 3 - 1 to it. At the same time the normaI calIs of kochi 1 Swartpoort 6 -

0 and garrulus could be heard at a distance along the 2 P. carpi d river-bed. A similar observation was made near the e t Swartbank Weather Station. In its immediate sur- Gobabeb Up to a d roundings this slow, monotonous call with up to 16 N. of river 16? - (

r clicks was repeatedly heard, while from a distance, e P. hchi h near the river, kocl8i called. Attempts to record this s i 11- l call, and to make observations of carpi while calling, Gobabeb b S. of river 12 0.134 sec. 16 clicks 1.78 sec, u have failed up to now. P e h t

y SUMMARY b ACKNOWLEDGEMENTS d e 1) The call of Ptm-, the mast vociferous: of the t n southern African geckos,is studied. As the observations contributing to this paper a r were made over a period of sweral years and on g species, and 2) Each of the three gamlus, m~pb e various field-Mps, I benefitted from the kind assist- c kochi, has a different call. n ance of a number of persons who are too num- e c 3) The males of garrulm and kuchf have been erous to be mentioned individual1y. However, I i l observed and photographed while calling and r would like to mention some in particular. In the first e their behaviour is essentially the same. P. carpi d place, I would like to thank Dr. V. F. FikSimons n is rarer and no definite observations have been for his patience and willingness to help and advise u made yet. y me. His encouragement and support, as Director of a 4) Certain populations of gams~luahave more than

w the Transvaal Museum, enabled me to undertake e t one type of call. many f ield-trips. a 5) The call of garmlus shows extensive inter- G am 0. M. I indebted to Mr. P. Prozesky for his t population variation and cannot be correlated e keen interest and the invaluable tape recordings he n i with morphoIogical variation. made on my behalf; Dr. J. C. Poynton for heIp b a 6) The pulse repetition rate of the call and the with the initial sound analysis; Professor P. de V. S

y vocal activity within a population of gamlus Pienaar and Mr. I. S. Hey of the University of b d e c u d o r p e R SCIENT. PAP. NAMJB DESERT RES. STN, NO. 46 93

~retoriafor assistance with the sonagrams; Pro- Davis, D. H. S. 11948). Sylvatic plague in South Africa: fessor Don Hunsaker fl from the San Dicgo State History of plague in man, 191943. - Ann. CdIege, California, for valuable discussions in con- trop. Med. Prrmsit. 42: 202-217. nection with sound production in ; Dr. C. K. FitzSimans, V. F. 11943). The Lizards of South Africa. Brain for reading the manuscript and helpfuI - Tu7. MKT, Men~oir,No. 1. suggestions; Mr. H. D. Brown of the Department of Goin, C. S. and 0. B. Gain 41962). Introduction to Herpetology. - W. H. Freeman and Company. Agriculture, Pretoria, for inviting me on three San Francisco. collecting trips; Dr. C. Koch, Mr. J. du Preez and Haacke, W. D. (19W).Description of Two New Species Mr. E. le Riche for information regarding the of Lizards and Notes on Pilzvimotrsin ba-wtpes seasonal activity of Ptenopus; Miss E. Olivier for (FitzSimons) fmm the Central Namib Desert. preparing one of the diagrams and Mrs. M. Erasmus - Sciewt. Prrp. Namib Deserf Res. Stli.. No. 25: for typing the manuscript. Sincere thanks are also 1-15, 6 pls., 3 figs. due to the late Director of the Transvaal Museum Hunsaker, Don 11, J. Strauss, A. C. Norton, G. A. Mish- Dr. W,J. Steyn, for his support of this work, as well ray 11068). Physiology of Audition and Vocaliza- tion in Lizards. Unpublished manuscript. as to my wife, for help in preparing this paper. Kelcman, G. (1963). Comparative Anatomy and Per- formance of the Vocal Organ in Vertebrates, in 'Amu~ticBehaviour in Anin~als',edited by R. G. Busnel. - Elsevier Publ. Co., Amsterdam - REFERENCES London - New York. Loveridge, A. (1947). A Revision of the African Lizards Andrieu, A. J. 11963). Techniques used for the Physical of the Family Gekkonidae. Bull. M7cs. romp. 2001. Study of Acoustic Signals of Origin, in 98, (1): 1469, 7 pls. 'Acoustic Behnvlour of Animab', edited by R. G. Mertens, R. 11946). Die Warn- und Droh-Reaktionen Busnel. - EIsevier Publ. Co. Amsterdam - dcr Reptilien. Abh. ~etickenbelg.~ttrtirrf. Gea. 471: London - New York. 1-108r 8 PIS. Brain, C. K. (1962). A Review of the Gecko Genus Mertens, R. (1955). Die Amphibien und Reptillen Siid- Ptmopzas with the Description of a New Species. westafrikas. Abh. se?rcke?iberg. nat~crf.Ge.7, 490: - Cintbebrrsla, No. l: 1-18, I pl., 2 figs. 1-172, 24 pls., 1 map. Broughton, W. B. (1963). Method in Bioamustic Termi- Smith, A. 11838/49). Ilkistrratio~lsof the zoo log?^ of South nology, in 'Acor~ticBehrr~~our of ', edited Africn: Rsptilin. - London. )

1 by R. G. Busnel. - Elsevier Publ. Co. Amsterdam Temhrock, G. (1959). Tiers!ir?inten. Eintr Eillfiihrung ill 1 - London - New York. die Biaaklrmlik. A. Ziemsen Verlag Wittenberg.

0 - 2 d e t a d ( r e h s i l b u P e h t y b d e t n a r g e c n e c i l r e d n u y a w e t a G t e n i b a S y b d e c u d o r p e R