300 i: 
 
 . . LTOH 
 
 Approved: 
 
 V 
 
 $LLT&f***r~Izr- 
 

 * \ 
 
 - 
 
Contents 
 
 I* Introduction 
 
 II Acknowledgments 
 
 III Materials 
 
 IV Physiological states 
 
 1. Rigidity and attachment 
 a. Persistence for -weeks 
 
 2. Locomotor 
 
 3. Resting or "active unoriented" state 
 
 4. Comparison of the physiological states of the different 
 species studied* 
 
 V. Responses of a single tube foot 
 1* Extension. 
 
 a. Conditions of extension 
 
 b. Direction of extension 
 
 1. Locomotor starfish 
 
 2. Stationary starfish 
 
 3. Rigid starfish 
 o* Mechanism 
 
 1. Normal 
 
 2* In isolated tube feet 
 
 3* Mechanical curvature, 
 
 4. Physiological inertia or lag 
 2. Attaching 
 
 a* Conditions 
 
 1* Physiological state 
 b. Strength of attachment 
 c Structures involved in attaching 
 
 1. Attaching reactions of isolated tube feet. 
 
 728925 
 

 
d. "Dependence of attaching reaction in isolated tub* faet 
 
 upon physiological state of organism, 
 . Attaching by only a part of the ambulacra! disk, 
 3* Releasing of attahment and withdrawal 
 a* Result of stimulation of side of column* 
 b. Response to st&sulation of the disk, 
 o. Detaching and withdrawal of isolated tube feet, 
 d* Detaching and withdrawal as a response to physiological 
 
 conditions in the nervous system* 
 4* The step reflex* 
 
 iJjW* 
 
 a. Intergradations with drawing response* 
 1 Difference in extension 
 
 2. Difference in withdrawing* 
 
 b. Description of the step reflex 
 
 0* Significance of extension^ during back sweep* 
 
 d* Analysis of the contact stimulus which initiates the 
 
 step reflex* 
 Analysis of the factors asrerning the orientation of 
 
 the step reflex* 
 
 1* Extension determined by "physiological an^terior* of 
 animal 
 
 2* Lashing back determined either 
 
 a* by location of contact stimulus or 
 
 b* tho condition of relative excitability of the different 
 
 parts of the longitudinal musculature* 
 statue of the attaching relfex during the step reflex and its 
 
 modifications* 
 
 g, Delation of the attaching reflex to the amount of resistance 
 to the step* 
 1* Methods of studying* 
 2, Numerical expression of this relationship Asterina 2*7 
 
Pycnopodia 2.06 
 
 3* This is no expression of the relative attaching 
 
 ability of these animals when not in the locoiaotor state, 
 h. Strength of the step reflex (pulling ability) 
 
 1. Pulling ability in Pi skater ooraoeus 
 
 2. Pulling ability in Ast 3rina.Miniata 
 
 1. On a solid substrate with and without additional 
 weight on its dorsal side. 
 
 2. On sand with and without load 
 
 3* Pulling ability in Pyonopodia helianthoidas 
 
 1. On said substrate with and without load 
 
 2. on sand with and without load. 
 
 VI. Coordination of the tube fast. 
 1 Preliminary description 
 
 2. Coordination in the tube feet of the rigid starfish 
 
 a. Retraction 
 
 b. Extension 
 
 c. Nervous mechanism 
 
 3. Coordination in the gills 
 
 a. Ciliary currents in gills. 
 
 4. Coordination that involves some orientation of the tube feet. 
 
 a. Coordination to attached tube foot and step reflex. 
 
 5. Coordination to passive movements of tube feet. 
 
 a* After twisting tube foot. 
 6* Coordination of the tube feet in active starfish. 
 
 4. Tendency of each arm to migrate in its own direction. 
 
 VII. Formation of the unified impulse* 
 
 1. General statement of the mechanism of the positive response 
 
 2. General description of the negative response. 
 
 3. Detailed description of positive and negative responses in 
 Pyonopodia. 
 

 
4* Orientation as a result of stimulating the dermal nei 
 net, or a general stimulation of all the tube feet* 
 
 6* The significance of the negative behavior of the isolated 
 
 rays* 
 VIII* Behavior of the starfish when under the influence of the 
 
 unified impulse 
 
 1* Positive reaction to contact and other stimuli* 
 3* Negative reaction to contact and other stimuli 
 3. Physical as distinguished from physiological orientation* 
 a. Direct orientation of the tube feet of the leading ray 
 
 from unilateral stimulation, 
 b* Acceleration cf the lateral rays by stimulation or fy 
 
 mechanical factors* 
 
 o* Retardation of the lateral rays by stimulation or -^ 
 mechanical factors* 
 
 XX* General oonsiderationof coordination* 
 1* 
 
 General consideration of the factors involved in governing 
 
 the direction of locomotion in the starfish and their 
 very delicately inter -related balance* 
 2* Theories of the mechanism of coordination* 
 3* Orientation of retracted tube feet and the independence 
 of the mechanism of orientation, and that of withdrawal 
 or stepping* 
 
 X* The breaking up of the coordinated impulse into areas in whio h 
 the tube feet are oriented in different directions* 
 1. The adaptiveness of this response as illustrated in 
 a* 'Ihe righting reaction 
 b* The teviation reaction 
 
 o* 'he locomotor starfish with curved lateral arm* 
 2* Physiological explanation not to be found in fcypothetioal 
 "Complex coordination center". 
 

 
 A 
 
3. Possible physiological explanation in the traction on the 
 tube feat resulting frora the movement of the rays over th 
 substrata. 
 
 a, Application of this to Uangold*s starfish, to the right- 
 ing starfish and to the deviating tarfish. 
 
 b. Svidenoe that the traction of the substrate does orient 
 the tubo feet. 
 
 1. Direct evidence inconclusive. 
 
 2. evidence from neurotoiaized animals* 
 
 5. Evidence from the behavior of the animal when its parts 
 
 are placed on separate substrates. 
 4* Evidence from the deviation reaction. 
 
 1. Deviation reaction not interfered -.vith by cutting nervous 
 connections *ith int^r-radial area. 
 
 2. Deviation reaction not elicited by prodding inter- 
 red ial area, 
 
 3. Quantitative aspects of the "deviation rush* with 
 different weights on the animal vary irith mechanical 
 conditions while quantitative aspects of stimuli re- 
 quired to initiate the negative reaction do not* 
 
 4. Operation of a tendency to return to original direction. 
 XI. Coordination of mov3mants of the tube fet with those of the 
 
 arm as a whole* 
 1* Illustrations of the tendency of an arm to set itself more 
 
 nearly at right angles to its actively oriented tubo feet, 
 
 when such movements involve dorsal and ventral flexion and 
 
 lateral twisting. 
 2. Ventral flexion of riid, of injured, and of nicotinized 
 
 starfish* 
 

3, :3esoripti.on of various other correlated movement* of 
 the tube feet and arms* 
 
 * 
 
 4 Description of th formation of the coordinated irapuliie 
 then the tuba fset ara free of the substrate* 
 
 5* Correlation of thaw Movements with the righting reaction, 
 a* Analyses of Jennings seven types of righting reactions. 
 
 6* Description of th righting reaction as it occurs vhan 
 the tube feet are prevented attaching by inverting the 
 animal on sand* 
 
 XIX Interpretation of the righting reaction as a phase of loco* 
 motion* 
 
 1* 'Yidenae from the movements of the tube feat and arms* 
 
 :*videnae from the fact that stimulation of the dorsal 
 
 rnyodenaal sheath is not mn essential faotor in the right- 
 ing reaction, 
 
 Z. .Tidenoe from tlie pereistsnoe of the 'unified impulse* 
 in the oam direction, to <* degree quantitatively compar- 
 able to its persistence in ordinary loccuaotion 
 
 

 
IH3RODU3TIQH 
 
 Although the behavior and physiology of starfish and 
 other aohinoderas have been given the attention of many and eminent 
 naturalists, it was hoped that an intensive study of the problem of 
 coordination in the several speoies available would bring to light 
 some data, that might prove of interest to the physiologist and 
 general aoologist. 
 
 The work was commanded in the autumn of 1917, but in 
 December was interrupted by fourteen month* s service in the army* 
 Between February 1919 and June 1920 I hive spent most of my free 
 time experimenting upon and observing the activities of starfish* 
 It would be quite impossible to set down my data in full, following 
 each experiment and observation out in detail, for reasons of space 
 alone* isy evidence, therefore, has undergone a rather severe select- 
 ive process* 
 
 I wish here to express my thanks to Professor s, J* Holmes 
 under whose direction the following study has been made, for his 
 careful criticism and his many helpful suggestions* I am greatly 
 indebted to Professor W* K* Fisher, of the Hopkins Marine Station 
 of Stanford University, for his courtesy in putting the facilities of 
 his laboratory at my disposal, and for bis help in collecting and 
 keeping alive the material. He was also kind enough to detenoine the 
 speoies I worked upon* I wish also to express my thanks to Professors 
 S. S, Maxwell and T. 0* Burnett, of the Physiology department of the 
 University of California for their helpful advice* 
 

 - 'At fff 
 
 , . K>:-> 
 
 viijwsii 
 
 
 
-2- 
 
 To ay wife, for her many cheerful sacrifices and her will 
 ing help in numerous ways, is due my fullest gratitude* 
 
 The following starfish were studied intensively: 
 Pi ea, star oohraoaus (Brandt) 
 
 (Stimpsonj 
 
 Asterina miniata (Brandt) 
 
 Supplementary observations ware made on the following 
 eohinoderms: 
 
 I^ptasterias eaualia (Stimpson) 
 Pisaster bravigpinua (Stiiapson) 
 Bvagterias troachelii (Stirapson) 
 Stromgrooentrotua franoiaoamua (Agassis ) 
 Professor 9* K* Fisher writes me as follows "Jennings 
 (1917) worked on Asterias aertulifera Xantus* I hav the actual 
 specimen sent, for identification to the Museum of Comparative 
 Zoology. Verrill calls the sane species Qrthasteriae gonalena." 
 Jennings uses the name Astarias forreri De Loriol* 
 
 So far as X am aware, the above seven species are the only 
 Pacific coast starfish, whose physiology has been described. 
 
 BUL.-w-_*-*y T .-.T-^g??g 5*. ^ *V -JMHLO. J* 
 
 PWMM 
 
 Piaster ooraoeus, was collected from the wharves in 
 Oakland harbor for study in the zoological laboratory of the Univer- 
 sity of California. Jor study in the laboratory of the Hopkins 
 Marine Station they were obtained from the surf beaten rooks in 
 front of the building. 
 
 A remarkable difference was evident in the physiology 
 

 
 
 
 
 --JUttS 
 
 
of the specimens taken from those tiro locations, which was not, BO 
 far as I was able to determine, due to the salinity of the ;ater in 
 the aquaria, its temperature, freshness, air content or tha food needs 
 of the animal * 
 
 Pilaster taken from the surf -beaten rooks were vary inac- 
 tive, would attach tightly for Ions periods of time to the substrate* 
 and could not be exoitad to active loooiaotion by the most varied, per- 
 sistent, or continued stimulation* The water in the aquaria was run- 
 ning freely and would keep these animals alive and other animals 
 ( starfish, crabs, sea-urchins etc.,) alive and active indefinitely* 
 
 The specimens of Pisaster ooraoeua taken from Oakland 
 harbor presented, ^hen fresh, activity of an almost opposite nature* 
 It was quite as difficult to get them to stop crawling as it was to 
 get those from the surf beaten rooks to start* In some specimens this 
 state of extreme, activity never appeared} but in the large majority 
 it appeared -7hen the animals were first put in the aquarium and, con- 
 tinued, interrupted by rest periods of greater or less extent, for 
 from two hours to two months* 
 
 The only speoiaen from the surf -beaten rooks at Pacific 
 Grove which showod this marked looomotor activity was one that had 
 been in the quiet water of tha aquarium for nearly thre weeks* At 
 the end of this period the animal forsook the tight clinging which 
 had occupied it during its struggle to maintain a foot hold on the 
 rocks and began active migration* 
 
 The specimens occurring on piles in the relatively quiet 
 waters of Oakland harbor do not attach very tightly, though they can 
 do so when disturbed and are not nearly so prone to attach Then 
 brought to the aquarium* 
 
 X am not inclined to attribute this behavior to "learning" 
 

 
4- 
 
 (see Sterne 1891} nor even to habit formation (Jennings 1907), but 
 would explain it moro simply as a very marked and striking example 
 of "physiological inertia" t : (Jennings 1907) or tha tendency to 
 tontinue past responses in spite of present stimulations, v/e shall 
 inquire further into the nature of this tendency, (see also Romanes 
 & Swart (1881), Prayer (1886), Mangold (1908^) f Bonn (1908), Oowles 
 (1911 ), Holmes (1911), lole (19150). 
 
 To the tto physiological states above noted, the one of 
 extreme rigidity and attachment and the other of aotivs locomotion 
 with the arms more or less extend ad and flexible w may add a third 
 state in -vhich the arms are eat tended as in the loooootor st^te but 
 the tube foot are not oriented in any particular direction as they 
 are in the locomotor aninal. The tube feet are more or less aotiye 
 and not tightly attached. 
 
 Aniaals in these three states will be referred to as 
 (l) lodoiaotor or crawling starfish, (2) rigid starfish and (3) active 
 but unoriantodjOr resting starfish respectively* In these different 
 states the animal's behavior is wholly different* 
 
 Pyonopodia heliao trifles the large 20 rayed "sun star* 
 present these same physiological ststes in quite as marked a manner 
 as Pisagt-sr. I hive nevar observed pyonooodia to assua the rigid 
 or attached state fhen on a horizontal substrate* It will attach 
 quite readily to a vertical substrate, and with such tenacity that 
 it is very difficult to rajaove it, but on a horizontal substrata I 
 have observed it 
 

 
 
 
 
 
-5- 
 
 only in the looomoto^or resting (active but unoriented) 
 state. 
 
 In Asterini the physio logical states are not wall 
 differentiated. The animal dees not attach tightly though 
 it does become rather rigid and inactive. he looomotor 
 
 ^,yV>XwW 
 
 state is olearp^lthouEh in the unoriented state one often 
 
 tb 
 sees the snimil make lurches, as if 4&e orawl in this and then 
 
 in that direction without actually doing so* 
 
 The other starfish observed seem to present different 
 
 physiclp ioal states .-uore or less analogous to those described 
 
 for Pisaater. 
 
 In the folio-fine pages we shall discuss the responses 
 
 of the tube feet as individual organs, their coordination 
 
 among themselves, and the relation of these movements to the 
 
 c.-ordinition of locomotion and righting* 
 
 ft* 4 sipofrg xuas POOT 
 
 The tube foot of a nomnl starfish may exhibit the follow* 
 ing responses, .tiich vary, as we sJnll see, with the physio* 
 logical state of the aniaal: (1) extenaion ; (2) attaohing(S) 
 withdrawal, (4) step reflex* 
 
 3HTBHSION 
 
 -Extension of the tube feet is best seen in the active 
 starfish upon the absence of those mtimulations which normally 
 cause a withdrawal of the tube foot or complicate its extension 
 by inducing the activities of attaching or ""stepping." 
 
 70 study the factors which govern simple extension of the 
 tube feet it is necessary then to invert the animal on its 
 abofal side, or better yet to suspend it freely in the water. Thus 
 are avoided the disturbance of contact stimulation. 
 
6- 
 
 Direotipn oj[ extension 
 
 The extension is conditioned in direction by the looomotor 
 activity of the animal as a whole. If the starfish is/ uigra ting 
 in the direction f a certain arm, for instance, the tube feet 
 will, in the absence of contact stimulation extend themselves 
 in this direction, and remain so extended until stimulated 
 either to retract or execute the step reflex. 
 
 In the stationary, non-rigid starfish the tube feet of 
 the outer part of the ray are, in the absence of contact simulation, 
 extended more or less toward the tip of the ray and ant moving 
 (feeling") about in that direction* This of course is not 
 constant and is laost noticeable in the most active specimens. 
 
 Starfish that are inactive or in the rigid state do not 
 extend the tube feet as much as do individuals of the active 
 non-loooaotor type. Vhe most noticeable difference between the 
 behavior of the tafcaartocxxocfxtloi tube feet of such a starfish 
 
 and those of a normally active one is that the former are not 
 
 avyy 
 directed <n*t from the tips of the rays* They may be waving 
 
 about approximately at right angles to the ray or even direc- 
 ted somewhat toward the center. 
 
 The mechanism of extension, first described by Heamur 
 (1910) in a very interesting paper is well known. It involves 
 a contraction of the aabulacral ampulla and a relaxation of the 
 longitudinal musculature of the tube foot* To ascertain the 
 dependence of this relaxation of the longitudinal musculature 
 on the radial nervous system, tube fe ;t were cut off and tied 
 on to the end of a capillary glass tube* This was connected with 
 a column of sea -water arranged so that the pressure could be 
 

 
7- 
 
 i nor eased or decreased by raising or lowering a reservoir, which 
 was connected to the capillary tube by a lon& rubber tube* 
 If the tube feet were injected with water at a pressure of 
 10 cm (HgO) they would slowly extend in the absence of contact 
 stimulation but not to their .vhola normal length* The extension 
 was much slower than the active extension of a norraal tube 
 foot and not so complete. If caused to contract and then 
 injected *ith a pressure of more than 2 ra (HgO) the extension 
 was not appreciably accelerated but could be made more complete* 
 
 Tube feet anaesthetized in Kg SO A would extend completely under 
 
 * 
 
 low pressures. This anaesthetization involved/ also relaxation 
 of the circular muscles so tint the tube foot presented a 
 noticeably greater diameter than the normal tube foot* In the 
 extended as well as the contracted tube feet there was a quite 
 
 constant curvature in the direction of a clear longitudinal 
 
 <3l0n<?11ze side 
 Line up-^hrr-sfet of the pedicel which I take to be the 
 
 pseudohaemal canal (Ouenot 1888) This curvature persists in 
 the anaesthetized (or dead) pedicel and is therefore probably 
 due to mechanical rather than to physiological factors* 
 
 Active tube foot preparations were allowed to extend and 
 assume their normal curvature toward the pseudohaemal canal, 
 and then were bent slowly and gently in some other direction. 
 They showed a tendency to remain bent in that direction and 
 then slowly to bend back to the original curvature* An 
 
 anaesthetized or a dead tube foot does not show this behavior* 
 
 c, 
 It is hense physiological in its nature and is perhaps 
 
 analagous to the behavior of a aea-urohin's spine when bent 
 over to one side/ (Yon Uexkull 1900 ) 
 
 Attaching 
 
 Attaching is conditioned by the physiological state of 
 

 
 
 
8- 
 
 the organ! am* The tube feet of Pisastar in ordinary looomotion do 
 not attach very strongly* ./hen in the rigid appressed stAte, how* 
 ever, they are ao tightly adherent that many may be pulled off be* 
 fore the animal can be removed from the substrate. 
 
 Strength ojg attachment* 
 
 Mr. '.Vaymouth of the physiological department of Stanford 
 University informs me that he haa released the tube feet of suoh 
 a starfish one by one with a needle until there were just enough 
 tube feet adhering to suspend the Animal from the lover surface of 
 a glass plate* The estimated area of the disks of these tube feet 
 multiplied by atoosphoric praasure was approximately equal to the 
 weight of the starfish thus snoring that these organs are mechani- 
 cally quite efficient. 
 
 ^Structures involved in. attaching. 
 
 Attaching is a reflex *hioh, though it may be modified 
 by outside factors, involves neoossarily only the muscular and 
 narvous structures of the pedicel. 
 
 Tube foot preparations texre made as above from actively 
 
 attaching starfish, great oare being exercised to quickly and 
 
 A 
 
 gently. It *as found upon placing suoh a tube foot against a stub- 
 strate that in about five oases out of ten, it would attach and hold 
 against considerable tension (in one case enough to tear off a 
 part of the disk). This po^er of attaching was lost after a few 
 trials. 
 
 .o..f attachin roaotiQn jn iaolatsd tube feet 
 
 upon i^iyaiologioal .state o.f 
 
 Tube foot preparations were also made from starfishes that 
 were not attaching (in active looomotion, feeling about the surface 
 film, etc.) These did not attach. 
 

9- 
 
 The interpretation of this phenomenon is rather difficult. 
 It is well known that when aa attached starfish la pulled off from 
 its substrate, many of the tube feet will be torn off and may remain 
 attached to the substrate for some time* The experiment shows fur* 
 ther that mioh a tube foot nay reattaoh even tho it be unconnected with 
 the radial nervous system (sea also Botazzi 1898 and Ruaso 1913)* 
 
 It is well Vnown, also, that some times a a starfish is very 
 
 \ 
 
 prone to attach its tube feet tightly to the substrate while at other 
 times the animal's energy is taken up with loooiaotion or some other 
 activity that does not entail tight attachment of the tube feet 
 (Jennings 1907)* The experiment shows also that there is a difference 
 in the behavior of the isolated tube feet which corresponds to the 
 fluctuation of the attaching reaction of an animal from time to time* 
 
 According to Von Uexkull, the contraction of a muscle is 
 due to "Tonuft* which is metaphorically referred to as a fluid, that is 
 carried to the muscle through the nerves* Furthermore (1903) by out* 
 ting the nerve which has supplied this tonus, the "fluid* may be en- 
 trapped in the muscle, and the muscle remain contracted* Vhile this 
 
 (t^n^e4d i<*0*lr) 
 
 theory has not been vory widely accepted, some of its aspects are 
 
 /\ 
 
 partly congruent vith the behavior of isolated tube feet* 
 
 Tube foot preparations, however, that are capable of 
 attachment do not present any differences in appearance from those 
 that are not capable of attachment* Thus they are not influenced 
 by entrapped " tonus* in the sense of Von Uexkull because "tonus" 
 elicits contraction or tension in the muscles it affects and the 
 tube feet undjr observation did not seem to differ in this respect 
 from tube feet which would not attach to a substrate* In fact they 
 differed from tube feet taken from a starfish in active locomotion 
 only in being in such a state of physiological activity that the 
 
* 
 
 
 
 
 01 
 
-10- 
 
 attaohing reflex la the one t/iut contact stimulation alioita. 
 
 It -would seem, then, from the difference in behavior of 
 tube feet taken from animals in different physiological atatea that 
 this etato of specialized irritability ia a condition of the ambu- 
 lacral Us^ and ^hilo angendarsd, most probably, by influences pro- 
 ceeding fro^a the radial nervoua ayatam, is not dependent upon that 
 aystea for a rather limited continuance, 
 
 Attaching bjc, only a, P-^ rt g^ ifce aabuliorapl digl: t 
 The : .tt-x9hing raflex does not, naoessarily involve all of 
 the ambul-5orl disk, rh.e and of a gnall rod was placed on various 
 partn of tii3 lo'tar surface of a larsa actively attatiiiing pediol 
 The part in cont-act -7ith the end of t>io rod attached with aroat force, 
 auoh thit :in atteispt to withdraw the rod result M in pulling a portion 
 of th;j disk out of shape. A fina hook *a laid flat against the 
 
 of tub3 r -j thit 
 
 atroaiant ^*5.9 hoo.V-ahaj 
 
 * 
 
 iak in contact *.ith the in- 
 attached to the hook quite 
 
 strongly* In fact, any part of varioua di*kf ^faa found to attach 
 even to the point of a nesdls, -shan tM was applied gently enough. 
 Thea exparimanta were repeated upon Isolated tube foot preparat-iona 
 vith. the same result. 
 
 Tha disk as an attacliing laeehaniara, than, ioea not act as 
 a whole (Preyer 1086), but rather the inaupping oooura to'vard the 
 center of any properly stimulated ai-ea. 
 
 Re\ea a ing and^ vitSidraval -*s a result of 8 t L.iiaul xticm o/ 
 p,ide p^* colxam 
 
 Rele&ae of attaclunant and withdrawal are Vo reaponeee 
 

11- 
 
 that are closely analagous* If a starfish IB tipjhtly attached 
 to the side of the aquarium, to gat it off without injury to the 
 tube fget, one has but to stimulate the aides of the tube feet sharply 
 Trith the eds* of seme flat instrument that will slip undor the star- 
 fieh, This stimulation causes the release of the stimulated tube feet 
 and sometimes the release of neighboring tube feat* 
 
 If a starfish be inverted or suspended, when not exhibit- 
 ing a locomotor tendency, and the side of an extended tube foot be 
 touched even v-jry lightly, thire ie an iflwoediate collapse and with* 
 draval of UXQ tube foot* Careful obpervation of the phenomenon 
 leads one to think that it is a reeult, first of the relaxation of 
 the ampulla and second of a contraction of the longitudinal muscula- 
 ture of the tube foot* 
 
 aif a, ra8onq9 J^o. glj^ulAtion o_f. 
 
 If the tube feet show a t'aidency neither to locomotion nor 
 to ttaohraent, this same withdrawal reaction follows the tianilation 
 of the diaV. 
 
 Usually, ho-vorer, Uiere is a tendency toward attachment 
 which does not necessarily interfere with th presence of the with- 
 drawing reaction* This conclusion was reached from a study of the 
 reactions of tube feet to very light suspended objects* A small 
 piece of thin celluloid, suspended by a thread, was brought in con- 
 tact with extended (non-locomotor) tuba feet* The first response* 
 usually was found to be attachment* After tide, dap ending on con- 
 ditions which will be 
 
 

 till 
 
12- 
 
 discussed in connection ith the step reflex, a 
 extension sometimes occurred due probably to an increased 
 tension of the ampullar muscles. Next/ in sequence in the 
 non-loco -otor tube feat was the retraction of the tube foot 
 and a consequent mo Ting of the piece of celluloid toward 
 the ray* Thin does not involve release of the substrate by 
 the disc as does the withdrawing on stimulation of the side 
 of the column yind is probably the response of the tube feet 
 t *t is involved when the ani.rul shrinks^ down on the substrate 
 after having ben disturbed* 
 
 and wi thdra'al oj; isolated tubft 
 
 An isolated tube foot preparation does not show typical 
 withdrawal reactions, because of course, the reciprocal action 
 of the ampulla is absent* Harsh stimulation of the column of 
 the attached tube foot preparation was fount to cause release* 
 Shortening by a slow contraction of the longitudinal 
 nusculnture was found to follow sever* stimulation f any part 
 of the tube foot, even against a strong water pressure , 
 
 Response jfco internal changes 
 
 Release and withdrawal nc of attached tube feet may occur 
 as a response to a change of internal physiological conditions* 
 Thus tin animal all of whose feet were tightly attached, one 
 minute, may the next minute be seen in active locomotion about 
 the aquarium* The factors governing this response will be 
 taken up elsewhere* 
 
 with withdrawing reaponsa* 
 
 The step reflex * is I think, merely a notification of 
 
 The first description that "l can find o'^ 1 the*steprefiex*" 
 is that Given by Reamur &71CJ, After describing the morpho- 
 logical connection of the ampullae ("tiny pearl like* balls') and 
 the "legs" (tube feat? he goes on to say "But one brings out 
 the whole ingenious mechanism of it when one presses the finger 
 on one of the MMlls**^ It is seen to empty and at the same 
 

 : " 
 
 
-US- 
 time, the 'leg* which corresponds to it becomes inflated and elonga- 
 ted, Finally it is aeon that on cessation of the pressure the ball* 
 refill and the legs become empty and shorten themselves, and it is 
 nothing more than this that the starfish does in extending its lega- 
 to press upon the balls, aa one may do at any time --vith his fingsr* 
 It is easy to imagine a thousand ways in which the starfish can do 
 this. The compressed balls discharge their -tater into the legs which 
 they inflate and thus extend, but when the starfish ceases to press 
 on the balls, the natural elasticity of the legs, -*hlch is consider* 
 able causes them to shorten* These legs, thus elongated the animal 
 uses in locomotion by t*.****^ extending them out toward the body to 
 which traa animal wishes to move and attaching to it at a vary acute 
 angle* The strength -vith which the leg remains affixed to this body 
 while trying to make a right angle -sith this same surface obligee 
 the animal to approach*" 
 
 of the withdrawing reflex as a responsa to contact stimulation of 
 the disk* The intargrading steps depend upon the presence to a 
 greater or less degree of a looozcotor tendency* This expresses it- 
 self, in the inverted or suspended starfish, as already shown by an 
 orientation of the extended tube feet in the direction of the phys- 
 iological anterior* If the locouotor impulse is not very strong, 
 the only modification perhaps that will toe observable in the with- 
 drawing reaction, will be an exaggeration of the tendency to extend 
 after the contact stimulation and before the withdrawal* 
 
 With the increase of the locomotor impulse comes a change 
 in the behavior of the tube foot wMoh integrates both with the with- 
 drawing response and the step reflex* this change is a further in* 
 crease in the above noted tendency to extend, caused no doubt by an 
 increase in the tension of the ampullar muscles* This complicates 
 the withdrawing action, and then results, for reasons which we will 
 take up later a more rapid contraction of the muscles on one side of 
 the pedicel than on the other* This gives rise to a lateral movement 
 of the tube foot 'thich increases in extent with the increase of the 
 locoraotor impulse, from a slight bending (fig* 3) of the tube foot 
 to one side, to an active lashing back (fig* 4) of the disk with 
 sufficient force to throw a grain of sand some few centimeters* 
 

 
 .** *e 
 
 
 
13ft- 
 Description of the step reflex. 
 
 Under ordinary cireunstanoes of locomotion, this lateral 
 oreaeat is followed by retraction and toe retraction by re*extenaion 
 in the direction of locomotion. This infolres contact vith the sub- 
 strata and UM itiaulations -*hich gire rise to the repetition of the 
 lashing baek, the retraction and the re*ztension These aoTeaents 
 vhioh inrolre, as sho-vn in detail later, attaohnent to tha substrate, 
 are tl'x>se of ordinary locomotion, iadi tube foot, acting indepeadeat- 
 ly as to tiae but in hansony with its felloes as to direction, re* 
 peats these ao Yemenis as long as contact stiaculi result H usi extension 
 and the locoaotor iapulse reaain unimpaired* 
 

14- 
 
 
 Si on 
 
 pciri'v,: . 
 
 ,*. V S V"" p- W ' 
 
 Jenr >*Tp. 90) desori *es the stp reflex in 
 
 torus of tha behavior of a tube foot on a solid flat substrate. 
 
 i > 5 4- U 7 
 
 9 ixy>g< 4^c {6*ftbing an arc, 
 as it does IBM stin* obja ,, light (See also 
 
 Ifameold 19084* It ia this tanftsmiy oo-tieaaribe an arc, to 
 keap fully extended as the disc is pushed back, that koaps the 
 animal veil off the substrate during locomotion, 
 
 A further analysis of the step reflex raises the question? 
 
, 
 
-15* 
 
 (1) What is the stimulus which sets if off? (2) what factors 
 govern its orientation? (3) What is the Status of the 
 attaching reflex in the various stages of its accomplishment. 
 (4) What is the relative strength of the otep reflex in 
 different species* 
 
 The stimulus 
 
 The stimulus which sets off the step reflex is one of 
 gentle contact on the disc, contact on the column or harsh 
 stimulation of the disc results in a simple withdrawal. In 
 absence of contact stimulation, there is no approach toward 
 the ste/p reflex. I have seen a large ffyonopodia on its back 
 in shallow water, remain with a large part of its 22,000 
 (Verrill 1914) tube feet extended in one direction (the direc- 
 tion changing from time to tine) for half an hour, with 
 
 ' wjP * ^^r i*f 
 
 none of the tube feet executing the step reflex* When, however 
 a light object, tmch a piece of celluloid was plaoed on the 
 tube feet the step reflex immediately started in all of the 
 tube feet receiving the contact stimulation* As a result the 
 piece of celluloid was quickly "walked* to the temporary 
 posterior of the starfish* The same was repeated with a very 
 thin clear glass watch-crystal* The glass could not be seen 
 at all, under water, but its course across the tube feet could 
 be clearly followed by observing the area in which the pedicels 
 were executing the step-reflex* 
 
 Ihen a starfish in aotive locomotion is brought above 
 the surface of the water the step reflex was seen to occur 
 without further stimulus* An aotive specimen of Pyonopodia 
 with the ventral side exposed to the air, presents 
 the likeness of some strange sort of military activity. lth 
 m*chinf like regularity the 22,000 bright yellow tube feet 
 

 
 
-16 
 
 ex tend themselves out toward the temporary anterior and 
 then lash back vigorously in the opposite direction, exactly 
 parallel with e-ich other* 
 
 The true significance of this is aean if the tube feet 
 of a part of such a starfish be submerged. Then only those 
 tube feet that touoh the surface film of the water, or those 
 entirely exposed to the air execute the step reflem. The 
 submerged tube feet remain pointed in the direction of the 
 temporary anterior until .some- contact stimulation, from the 
 surface film or from some solid object initiates the 
 step refles. 
 
 ^hat factors govern tiie orientation oJT the, step reflex? 
 
 The first phase of the reaction, the extension of the 
 tube foot is a function of the physiological orientation of 
 the starfish* This will be analy ed further elsewhere* Now 
 if the lashing bac> is to be effective in locomotion, it must 
 take place (as it does) in the opposite direction from the 
 extension. This, however, merely shows that the response is 
 adaptive and is not a physiological explanation* A physiological 
 explanation may be looted for in the location of the oontaot 
 stimulation on the disc of the tube foot or in the condition 
 of tension in the musculature of the column* The tube foot 
 as it extends may be seen often* tho^not always * to touch 
 the substrate first with the poi^fMMtflfe* ** mi y be 
 
 X. 
 
 
 expected that exitation to oonw^w. _ ontaot 
 
 stimulus might spread to the side of the column 1*3 and cause 
 its contraction more quickly than to the side 2-4* Furthermore 
 a contact stimulus at the place 2 does not elicit the step 
 

-17- 
 
 reflex with as much readiness and regularity (Pisaster) as 
 similar stimulus at tho place 1. It must be remembered 
 however that in normal locomotion the diso is often placed 
 flat on the substrate, tid that when the tube feet are exposed 
 to the air the ourfaoo tension fil.-i may be expected to oontnot 
 with equal pressure on all sides of the disc, and thus to 
 stimulate them all equally* We ha TO to count then upon the 
 greater excitability of the muscles on the side 1-3 in the 
 post-contact phase of the step reflex* This is comparable to 
 the increased tension of the muscles on the aide 2-4 in the 
 pro-contact stage of the step reflex* The oscillations of the 
 tube feet may be explained in terms of Von Uexkiuflfs law of 
 "tonus" or may be left unexplained* the fact is, of course 
 that they move back and forth in the step-reflex with con- 
 siderable regularity and precise orientation* The factors thai 
 control the orient ition of the animal will be taken up in 
 connection with an analysis of coordination aoong the tube feet* 
 
 Status pjf foe attaohjlnf reflex during the stop reflex* 
 
 The strength of attachment during the atop reflex differs 
 as we shall see with the different species and with the amount 
 of resistance there is to the accomplishment of the step* 
 
 In general wo may assume from observations n ordinary 
 locomotion that the tendency to attach is strongest, during the 
 progress of the step reflex. Just after the contact* The 
 tube foot usually remains attached during the first half of the 
 backward oscillation, but the likelihood of release (or slipping) 
 is found gradually to onorease during tho last phase of the 
 atep reflex* 
 
 A largo grain of sand was placed on one of the ambulacra! 
 disos of an active Pyonooodia. The step reflex which resulted 
 was so violent that tho grain of sand was thrown as from a 
 

-18- 
 
 miniature catapult, a distance of four or five am. on 
 repeating this, the elevation or "angle of fire* was aeen to 
 be such as would entail release of the grain from the diao 
 during the third quarter of the arc that the diso describes 
 in lashing back. Usually, however, in disaster. Asterina etc., 
 the violence of the lashing back is not so sratt, and the 
 release is not very sudden or prompt eo that such a catapulting 
 action is not often seen in these forms* 
 
 a: refj.e^ tft tjhe, aPVffMt SLL 
 
 resistance ^o, Jjhe step* 
 
 The relation of the attaching reflex to the amount of 
 resistance to the step was obtained in the following manner* 
 One of the rays of an Agjfcer^na. was tied by a long thread 
 to a spring recorder which was calibrated to grams and set 
 to writhe on a slowly moving drum. Ihen the animal pulled 
 against the spring, the strength of the pull was recorded as 
 the height of the curve above the base line. Now when the 
 animal had pulled the spring up to various heights, the glass 
 plate on which it was walking was suddenly slid forward - 
 In the direction of locomotion* This resulted in an increased 
 tension on the starfish which was recorded on the drum until 
 this tension became sufficient to cause the animal to release 
 hold on the substrate* The curves got by this method were 
 somewhat as follows: 
 
 1-: the fctfMT' given by the starfish as it walka against 
 tht resistant j>f the spring. s was slid 
 
3SJ* 
 
-19- 
 
 forward and the curve 2-3 meaouros the amount of increased pull that 
 the starfish was able to resist before releasing (at 3). 
 
 The values for 12 observation on Asterim are as follows!: 
 
 Strength of pull 
 (2 on fig.) 
 
 2 g 
 3 
 
 5 
 6 
 9 
 
 12 
 
 18 
 
 18 18 
 
 18 
 
 27 
 
 33 
 
 Releases 
 at (3 on fig.) 
 
 15 g 
 
 15 
 
 27 
 
 21 
 
 24 
 
 36 
 
 45 
 
 3/2 
 
 7.5 
 
 5 
 
 5.5 
 
 3.5 
 
 2.6 
 
 3 
 
 57 
 60 
 66 
 84 
 
 54 aT 
 
 2.5 
 2.5 
 
 Disregarding the high values of the first three observations 
 due observably to the fact that certain of the tube feet were "re- 
 fractory", -that is, had not become coordinated in the step reflex and 
 wers simply attaching, -79 find that the strength of attachment of a 
 tube foot is on the average 2.7 times the amount of pulling the tubs 
 foot is doing at that time (amount of resistance to the step). That is 
 to say, the tube feat are attached strongly enough to resist a pull 
 ab-n.t 2.7 times as graat as that to which they actually are subjecting 
 themselves; a facto? of safety against skidding on the smoothest surface 
 of 2,7. The valu* of friction in the above experiment was tested with 
 the starfish inverted and found to be negligible (about 3 g). 
 
 'Whether the relation (quotient 3/2) between the t-/o variables 
 is constant, logarithmic or of some othar nature can be told only 
 after much statistical compilation of data. In Ajtarina it seams to 
 

 
 
 -fit* 
 
 
 
 
 
-20- 
 
 be fairly constant within the limits studied* 
 
 In ffyonopodia the relationship is even more constant, 
 though it has a -wholly diffe^nt value as seen from tha following 
 table: : 
 
 Strength of pull Release at 3/53 
 
 (2 in fig.) (3 in fig.) 
 
 9 
 
 18 
 
 
 2 
 
 18 
 
 33 
 
 
 1.8 
 
 24 
 
 30 
 
 60 
 
 A 
 
 
 2.5 
 
 
 36 
 
 72 
 
 
 2 
 
 Here the average quotient is 2. 06. The tube foot is 2,06 times at 
 strong to hold as it is to pull. 
 
 The difference in tha valu of vlie figure ia due to spec- 
 ific differences between the two starfish. It is not in any way 
 correlated with ability of the tube feet to attach when not in th 
 locomotor state. An attached stationary Asterina is very easily re- 
 moved from the substrate and only once have I seen a tube foot torn 
 off in the process. On the other hand Pyono podia the attachment 
 of whose tube feet during the st*p reflex is much less than that of 
 Aaterina. would when in the stationary clinging state hold with such 
 tenacity to the substrate, that it was only with much patience and 
 the loss of many of the animals tuba feet that I could pull it loose. 
 When the starfish was once released from the substrate, if the ten- 
 dency to attach continued, as it often did, I was confronted with 
 the equally difficult and much more unpleasant task of releasing the 
 animal from my own hands. I have spent the best part of an hour dis- 
 entangling the twenty-two arms of an eighteen inch Pyppnop o iia_ from 
 myself and the side of the aquarium* 
 
- 
 
 
 
 
 
 'ub 
 
-21- 
 
 g the stop reflex (Pulling ability) 
 
 Hot only does the ratio of strength of attachment to strength 
 Of pull # vary between different species, but also the pulling ability 
 
 # Soheinmetz (1896) states that a starfish (Aetsrias gluoialis) is 
 able to exert a pull of 1360 g in opening a bivalve, to which pull the 
 bivalve gave way, under experimental conditions in short order* His 
 method of measuring the pull, however, was directed rather to measure 
 the strength of the attaching reflex because he recorded the pull that 
 caused a starfish to let loose its prey and not the puU which would 
 overcome a maximal contraction of the longitudinal musculature of the 
 tube feet* The amount of pull exerted by a tube foot, under conditions 
 of locomotion at least, is aa we have seen from one -half to one-third 
 of the strength of attachment at that moment* Soheinmetz in this 
 interesting paper also lists five ways in which the starfish has been 
 supposed to open Oysters; : (1) by taking the mollusc by surprise, (2) 
 by besetting the oyster 90 long that it would be compelled by hunger 
 and want of air to open. (3) by hypnotizing the molluscs, (4) by boring 
 through their shell, (5) by poisoning them, all of which he shows are 
 fallacious* Beamur (1710) quotes Aristotle and Pliny as attributing 
 to the starfish a body heat, by which it kills its prey, derived no 
 doubt by poetic analogy from the stare of heaven* He himself believed 
 tint the starfish pries open the oyster with its oral spines and sucks 
 out the meat with its laouth* 
 
 considered alone* Tor instance, a small specimen of Pisactar about 
 
 12 cm in diameter was attached one noon to the recording spring and 
 induced to pull against it* During the whole afternoon the tension 
 varied between 40 g and 60 g* The drum was removed and the animal left 
 tugging at the thread all night* The next morning it was pulling in 
 the same direction but had advanced slightly* The tension during that 
 whole day varied from 95 to 190 g* There was much activity of the 
 tube feet when the animal was going forward or being pulled back by the 
 spring* When the animal was holding stationary tube feet were seen to 
 be arrested in the various phases of the step reflex so that only a 
 portion of them were extended forward at such an angle that they could 
 pull the animal forward* Toward evening the pulling increased and 
 somewhere between seven and nine p*m* reached a peak of 
 

 
225 g* This came from a sudden increase of pulling as shown by 
 the curve and resulted in the arm breaking off where it was tied* 
 The animal had thus pulled steadily at a tension of from 60 to 
 225 g for a period of over 33 hours* Another specimen 18 cm in 
 diameter pulled 300 g when it was released for fear of breaking 
 the apparatus* 
 
 Correlated with the fleet that Asterina. never attaches as 
 tightly as does Piaster is the faot that it never pulls as hard* 
 A 10 oca Aaterina, registered pulls of 60, 77, 69, and 46 g* in 
 four successive trials* A smaller (8cm) but more active As te ring 
 pulled 90g* 2he peak of the curve would be reached after a 
 gradual ascent of about 20 minutes* de decline would last from 
 one to two hours* Both the decline in the height of the curve 
 and the fact that the pull did not last long, comparatively, are 
 perhaps, evidences of fatigue* 
 
 To test the role of the attaching reflex in this response, the 
 animal was put on sand and set to pulling in the same way* the 
 best pull it could record was ?i g* A 40 g* (weight in water) 
 Syracuse dish was laid on top of the animal* This increased its 
 pulling ability to 15 g. The adding of weight to Asterina or 
 Pisaatar when pulling on a solid substrate made no appreciable 
 difference in theAr puHling ability* 
 
 The case of Pyonopaj&ia # is different as we shall see later. 
 
 Soheinmetz ^1896) states that with respect to food taking, 
 starfish may be divided into two types, those that swallow their 
 food whole such as Astro pec ten and those that pull open the 
 bivalves on which they feed and digest them by extruding their 
 stomach and applying it to the soft parts of the mollusc* (Asterias) 
 Although fyonopodia is grouped in the loroipulate. with AsteriaS. 
 and has tube feet, inoontradlstino tion to those of Astro pec ten. 
 capable of tight attachment, it swallows its footjwnole, ejecting 
 the (indigested parts* Correlated perhaps with the fact that the 
 animal does not pull open its bivavle prey* as do most of the 
 other ?orcipulata, is the faot that under other conditions as well, 
 the tube feet, though the;, can tightly attach, do not ordinarily 
 do so when pulling, and consequently the animal can not pull very 
 hard* 
 

 
 
 
-22- 
 
 i different no e sUali- see late*. The animal studied in this 
 respect *as about 50 on in diameter, with, aooording to Verrill*s 
 estimate about 22,000 tube feet, each of whioh was extremely 
 aq tire. In water the animal weired only 50 g. but in air the 
 weight was estimated to be well over 1000 g. Suoh a starfish 
 when set to pulling against the recording lever pulled 54, 45, 
 30,60 g* jin four trials ( on different days). The time relations 
 were similar to those of ftetorinals pulling reaction (less than 
 half an hour of inoreasing tension and up to two hours of 
 declining tension), 
 
 The remarlcable faot that this large and active starfish should 
 not pull marly as hard as an 8 om ftstarina. or less than one fourth 
 as hard as a 12 am Pj.sa.ster. wae thought perhaps to be due to 
 failure of the attaching reaction during the step-reflex, to keep 
 the same relationship with the resistance to the step (pull) for 
 these higher values, whioh it has shown according to the above 
 table for lower levels* Some tube feet were seen to slip on 
 the glass as they performed the step reflex* Other tube feet 
 were seen to be in the*refraotory state* that is to be attached 
 tightly and to be showing no sign of the step reflex. This made 
 it impossible to get direct evidence as to the status of the 
 attaching reflex in the looomotor tube feet, as the "refractory" 
 tube feet caused the release to be abnormally high* 
 
 Besides direct observation of slipping tube feet, indirect 
 evidence that the lacfc of pull was due to failure of the 
 attaching reflex in the active tube feet, was furnished by Wti/wcj 
 *bala*iittg" the animal with 80 ga (weight under water) of Syracuse 
 dishes placed on its dorsal side. When so weighted down, the value 
 of the 54 g. pull was increased to 69 g. and the value of the 
 
 60 gtt pull was increased to 75 The increased pulling ability 
 was undoubtedly due to increased friction between the tube feet 
 

 
 
 * 
 
-23- 
 
 and the glass It also involved the trenching loose of a number of 
 refraotory tube feet. 
 
 On sand it was found that the animal could pull 15 gm 
 
 W'ltll 
 
 ( without load) and tfce a load of 80 gm 
 
 oould pull about '62 gm* 
 
 OUORPINAVK.^f OF TH;; TUB& g&ffi 
 
 Preliminary description. 
 
 When starfish were suspended and the tube feet at the end 
 of one of the rays brought in oontiot with some solid object, those 
 that touched it first ~?are usually observed to attach* Then the 
 neighboring tube feat orient -3d and extended themselves in the same 
 direction 13 the attached tube feet* If opportunity offered theae 
 other tube feot attached as did the first tube feet* 
 
 If no'* these tube feet are stimulatad sharply they retract 
 and the neighboring tube feet also retract (Roman* and >iwert 1881, 
 Prayer 1836, etc*,}* The wave of retraction passes down the stimula- 
 ted arm, and out the other arms along the line of the ambulacra! 
 nervous system. This is in accordance with the older observers, es- 
 pecially Preyer (1386)* They also stowed that if the nervous syatea 
 was cut <*t some point the above coordination would extend as far a 
 
 the cut and no farther* 
 
 r 
 Further than the fact that it rests in the ambulaoial 
 
 nervous system, the mechanism of this coordination is very obscure* 
 Physiologically, it is a fact attested so far as I am aware by all 
 of the workers on this phase of echinodezm physiology* one tube 
 foot seems to "imitate" in its activity th behavior of its neigh- 
 bora* In the following analysis of coordination in the tube feet 
 we shall inquire into its 
 

 
 
 
24* 
 
 characteristics la the rigid starfish, and compare it vith 
 the ooordination manifested by the gills* '** wtti ilso inquire 
 into coordination in tube feet of active but non-oriented starfish, 
 the building up of this ooordination into the Unified impulse, 
 the behavior of tha starfish under the influence of the unified 
 impulse and the breaking do-am of this unified impulse under 
 various normal and abnormal conditions, 
 
 Coordination i& the tube*feet g the rigid '.atjarfish, 
 When rigid speoimene of Pisaster were suspended or inverted 
 the tube feet, after their temporary retraction from the stimula* 
 tion of loosening, were found to extend more or less at right 
 angles to the body of the ray. There were subsequent movements 
 of the ray *hioh vill be considered later* Some of the tube feet 
 were then stimulated to retract* There was a wave of retraction 
 passing along the lines of the tube feet* This lessened in 
 
 & d'ld 
 
 intensity as it proceeded from its source, so that it *y no 
 reach the farthest tube feet. Later the tube feet WM again extend 
 the wave of extension pawling back in the reverse order so that 
 the tube feet stimulated to retraot and those nearest them will 
 be the last to re extend* 
 
 To account for this ooordination in retraction and extension 
 it is not necessary to hypothesise very complex conditions in the 
 nervous system at the base of the pedicels* 
 
 ^'.uenftt (1888) Ludwig and Hainan (1899) ; tfeyer (1916) eto.,X the 
 ambulaoral nervous system seems to be merely a condensation 
 of the nerve net that extends over the outside of the myoderral 
 heath* So far aa I am aware there is no morphological evidence 
 of synapses in the nervous system of starfiahos, though of course 
 the evidence on this question is far from complete* A simple, 
 nerve net will account for the above behavior* 
 

 
 
-25- 
 
 Jt has been Been that an isolated tube foot will not 
 contract or extend quite normally. Certain conditions then may 
 be said to exist in the nerve net at the base of the stimulated 
 tube foot, which affect the muscles of the pedicel and ampulla and 
 cause the normal withdrawal (or extension) of the tube foot. Now 
 in accord with the well known laws of transmission of excitation 
 in a nerve net (Parker 191^ these conditions may spread in any 
 direction (within the ambulacjal nvrvous system) and cause the 
 retraction or extension of other tube feet. We shall see, 
 elsewhere that no such simple condition will account for the 
 physiological orientation of the tube feet and their coordination 
 in locomotion. 
 
 Coordination ig. gills. 
 
 The physiology of movement in the gills is quite similar to 
 that of the tube feet in the rigid starfish. Although there is 
 
 lateral movement in each there is no orientation of these lateral 
 
 (fa 4-W^u*- 
 movements in any particular direction in the gills. A- stimulus 
 
 will cause the contraction of one group of the (dorsal) gills, 
 will be communicated to others near these and cause 
 
 their retraction (Jennings 1907). In this region the nerve 
 net is quite diffuse, so that the spread of the contraction may 
 
 be in any direction. The wave of r -extension usually takes Adm^cn 
 
 -t* 
 opposite atii-tfutiun rrum^that of contraction. It is centripetal 
 
 rather than centrifugal. If the wave of retraction is sufficiently 
 strong it may be communicated to the tube feet and involve their 
 retraction as well. The retraction of the tube feet does not 
 involve the retraction of the (ambulaoral) gills (De Moor & Chapeaux 
 1691) an evidence of polarity in the nerve net which suggests 
 something in the nature of a synapse. That part of the nerve net 
 whicfe extends up the sides of the long ambulacfal gills in 
 Pisaster also shows evidences of polar^ation similar/ to the 
 

 '' 4,ts.S.r 
 
 - 182 SEX 11 
 
-26- 
 
 polarity of sea anemone tentacle (Psrliser 191$) in that when 
 stimulated at the base or middle, th* musculature, especially 
 the circular musculature, below (proximal to) the locus of 
 stimulation contract^ while that above (distal) dt>e$ not contract. 
 If stimulated at the tip the whole tentacle contracts, the 
 circular musculature responding to a lesser stimulation than the 
 longitudinal. If out off at the base with scissors, the 
 edges of both the stump and the ablated piece adhere together 
 along the line of the out by means, seemingly, of a sticky 
 substance on or near the cut edges, so that the wound does not 
 open an ^pJJ&rKture to the exterior. The stumps of course 
 shrivel down in strong contraction. They are found, three days 
 later a little short but with the end healed over normally. 
 The excised gills show no sign of contraction, and the cut end 
 being sealed over as describedabove, the gill remains distended 
 by its enclosed watef like a miniature "sausage balloon" with 
 a trun^a^ed end. The contraction of the gill musculature is not 
 sufficient to collapse the gill against the resistance of the 
 closed end. If this end be teased open gently and then the tip 
 be stimulated collapse ensues immediately* 
 
 Ciliary currents in frills. 
 
 J 
 One of the gills, when thus removed was seen to en&ose 
 
 several clumps of amoebocytes or wandering cells. These made it 
 convenient to see the ciliary respiratory current which continued 
 uninterruptedly after the gill had been removed. The amoebocytes 
 moved up one side to the tip of the excised gill and down the 
 other side to the base* It took three or four seconds to complete 
 the circuit. 
 
 Coordination that involves some orientation of, the tube feet. 
 
 Having studied the coordination of the non-looomotor tube 
 
 if 
 feet and compared #*a-fcwith coordination of the gills we shall 
 

already seen, they will attach. This ia usually followed by 
 
 -27- 
 
 now take up coordination in the behavior of the tube feet during 
 their transition stages between the looomotor and the non- 
 looomotor state* 
 
 If a rigid starfish be suspended and some of the extended 
 tube feet be broughct in contact with a solid object, as we have 
 
 increased activity of the neighboring tube feet and if the 
 
 v 
 
 starfish is not too rigid, by their active bending & *fcxx We 
 
 
 
 toward the stimulated cre<x .. It is in this phase of their 
 behavior, that the beginning of the step reflex can be 
 elicited toy proper stimulation* 
 
 Coori j.natip^ tp passive movements of tub& feet* 
 If on such a starfish a long tube foot be brought in 
 contact with a small object, such afe a pencil point the disc will 
 attach. If np'?i, the pencil point be moved, with the tube foot 
 
 still adhere'ing so that the direction in which the tube foot iS 
 
 r 
 now pulled out is different from that in which it originally 
 
 extended itself, other tube feet will thencoordinate themselves, 
 not in the direction of the original extension of the stimulated 
 tube foot but rather in the direction to which it had been passively 
 moved* This tendency to coordinate thus, while very marked in 
 some animals, is of course apt not to show itself in starfish that ar 
 are very inactive or very rigid, and is apt also not to 
 
 at all, if there is a strongly laarked coordinated impulse in 
 some other direction. Out of thirty trials on starfish in 
 various physiological states there was well marked and active 
 coordination to passive movement in fifteen. 
 
 This coordination could also be brought about when the tube 
 f*t was twisted by turning the pencil a few times ia the hand 
 before pulling the tube foot over in its new direction. I could 
 

 
 
 
 
 
 
 
-28- 
 
 
 observed^ no dif farenoe in the -accuracy or promptness of the 
 coordination, I have even untwisted tho tube foot again, in its 
 new position, without either disturbing the attachment of the tube 
 foot or the coordination of its felloes* Heedless to say these manip- 
 ulations had to be done with extreme care to avoid stimulations wiiioh 
 might cause retraction* 
 
 Coordination f the tube feet jjnt the active starfish. 
 
 thus far wa have baen discussing coordination in the tube 
 feat of rigid non-looorootor animals* But when a very large number 
 of tube feet are seen in tha suspended specimen, pointing in one 
 
 a 
 
 direction in ^coordinated manner, one is apt to be dealing -vith a star* 
 fish in the active rather than in the rigid state* 
 
 If we suspend a starfish that is active, but not definitely 
 oriented and locomoting in any one direction, we find that the tube 
 feet at the tips and for a centime tar or no re toward the disk are 
 oriented and actively feeling out toward the Up* Proper stimulation 
 of the tube feet at the ends of these rays will elicit the step re* 
 flex in the direction of the tip of the ray* This would indicate that 
 each ray has a tendency to migrate in the direction it points* 
 
 4 
 
 Tand.e.noy. oj each ray to. raipr-atje , to ward jLJ^ .LkJL* 
 That each ray does tend to migrate away from the disk was 
 demonstrated by attaohing five glass tubes or shell vials, large enough 
 to acooaodato the ray, to five floats and presenting these simultaneous- 
 ly to the tips of each of the five rays, in such a way that they could 
 
 each walk nto one of the glass tubes and in so doing pull it back over 
 the ray* When the rays got to the end of the tubes they were seen either 
 to keep on in the same direction or reverse and back out, or part way 
 out* It was raally quite amusing to watch this suspended animal indus- 
 triously 
 

 
 
-29- 
 
 < >' 
 
 trying to walk in five different directions at once. 
 
 Auto tony 
 
 Another indication of this tendency is the faot that in 
 stale water or under the influance of ohbroform (Moore 1916) 
 a starfish is extremely susceptible to autoton^ Pisaster 
 
 seems much more susceptible to this reaction if the nervous 
 system has been injured in some part. As I have observed it, the 
 reaction consists in an exaggerated tendency in the tips of the 
 several rays to migrate in their own direction and a failure 
 of this tendency to effect an orientation rf the tube feet of the 
 rest of the animal in the way that will be seen below to be 
 usual in the normal starfish* This is due to a pathological 
 sluggishness in the action of the central part of the ambulacral 
 nervous system, as seen from the fact that the tube feet in that 
 region are comparatively inactive. The raye of a Piaaster atwzatf 
 U'-vvd-t/w^tri^v^^ ' -f autotomy present an elongated appearance. 
 The tube feet at the tip pull actively, each in the direction 
 of its own ray, so that after stretching somewhat the ray gives 
 way, usually at or near the base. 
 
 FORMATION OF THS UNIFIiiH) IMPULSE 
 
 / - , 
 
 o< 
 
 From such a picture as the above it may seem as A far call to 
 the unified behavior of the actively walking starfish. In the 
 latter each tube foot is put out in a single definite direction and 
 locomotion proceeds in a beautifully unified and coordinated 
 manner. The difference is ttm Just this, that in the unified 
 locomotor starfish, one, or more often two adjacent rays become 
 for some reason more active than the others and the coordinated 
 state which is present at their tips spteads maintaining its 
 own direction and gaining impetue, over the other rays. 
 
 It u>\\^ .. be our purpose now to inquire into the factors which 
 give precedence to the activity of some ray or rays in the 
 

 
-30- 
 formation of the "unified impulse". 
 
 The responses of a starfish to stimuli, in so far as they 
 involve locomotion, may be divided into two categories, positive 
 responses, in which the resulting locomotion is toward the 
 stimulus, and negative responses, in which the direction of 
 locomotion is aw->y from the stimulus. Gentle contact at the 
 ti of the ray will usually elicit a positive response while a 
 negative response usually results from severe prodding or pinching. 
 
 General statement of the mechanism oJT the positive response. 
 
 The mechanism of the positive responses, is as I see it 
 as follows? A gentle contact stimulation of the tube feet at the 
 end of a ray causes these tube feet to extend in the direction 
 of the stimulus as we have already seen, other tube feet behind 
 this coordinate in this action, and receiving the contact 
 
 .. i i ' i "Hi > ** 
 
 stimulation of the substrate execute the step reflex* The 
 impulse to coordinate with the active tube feet at the tip of the 
 stimulated ray this spreads to the rest of the starfish, involving 
 after a time every tube foot in the body in coordinated locomotion. 
 
 General description of the negative response. 
 
 The negative response is brought ab*at on exactly the 
 same principle. The prodding or pinching of a certain ray 
 results in the retraction or inactivation of the tube feet in 
 that region and aa> &p^3a@W)s&8& vo the spread of this impulse, 
 to certain of the other tube feet. The extent of the spread is 
 of course determined by the strength of the stimulus* 
 
 Assuming first that the stimulation is severe enough to 
 cause all the tube feet to retract or become inactive, H 
 
 the first tube feet to resume their normal function are 
 
 those farthest away from the source of stimulation. In this 
 experiment the tube feet farthest away are those of the opposite 
 ray tips. These tube feet are ori'rlejted in the direction of their 
 

 
-31- 
 
 which is in fact away from the source of stimulation* In so 
 doing they come in contact with the substrate and execute the 
 step reflex* From this point on ; the coordination completes 
 itself in the same manner as outlined for the positive response* 
 
 In case the stimulation it not sufficient to cause the 
 retraction or inaotivation of all the tube feet, it will spread 
 
 the tube feet, to a certain extent so that the farthest tube 
 feet are the most active and therefore will dominate in the 
 coordination* 
 
 Detailed description ojF positive aftd negative response in 
 Hrcnopodia. 
 
 FHono podia on account of its large size and great activity 
 is very favorable for a study of the mechanism of coordination 
 in positiw and negativ* -"-a-nonaes. The active but not oriented 
 
 ,qu 
 
 
 
 a the tube feet at the 
 
 tip oi Sited in the direction of the ray, and 
 
 r upon proper stimulation. Now 
 
 A I A 
 
 convinced me that a positive stimulation 
 
 f coordinated activity in the 
 
 C 
 
 ordinated activity in the way 
 

 
 
 . 
 
-32- 
 
 dlagraaed above^ Inthis way <pand d will be coordinated before 
 b though b and the neighboring raya ay be ;oore active in their 
 coordination than o and d because they receive stimulation 
 throu^i the rin : fro;.; both directions simultaneously. 
 
 Now with the negative # response, conditions are different 
 
 # The negative response has been described by Loeb (1900) 
 in terms cf observations by Norman (1900) as a result of the 
 retraction of tube feet on the harshly stimulated ray and a 
 consequent determination of the direction of the negative 
 responses by a "parallellogrim of forces" exerted by the other 
 rays, each, hypothetioally, as I take it, continuing, during the 
 negative rotation to pull in its ewn direction. It is -eil icnown 
 from the -/ork of l-tomanos, Preyer, Jannings, Mangold, Cole and 
 others tjiat all normal locomotion is brought about by the cooper- 
 ation of all of the tube feet stepping in one direction and not 
 the divergent pulls of the various ra s, which as w have seen 
 results in autotomy. 
 
 in certain respects. Assuming that the harsh stimulus is given 
 
 at b. The p*tk olLJEtttraaUoa JU1 w a* above ( but the Way 
 the coordination impulse spreads is tmtz identical with that 
 
 |X 
 
 diagrammed in fig. 2 so that o_ and d. become oocrdinated before 
 b, which is the location of the stimulation. 
 
 Appearances seem to indicate that just after the retraction 
 following such, a negative stimulation, the tube feet on the 
 far side of the animal show a definite increase in activity. 
 
J.SO 
 
 
 
 
Whether this increase is only relative or to what extent it is absolute 
 I am unable to say* 
 
 Eunotion o^ the qtej reflate jji the spread gf, coordination* 
 The function of the step reflex in the spread of coordination 
 ia probably very Jj.iocrtit The pinching of one ray of an Aet^rina will 
 cause prompt negative locomotion with all the tube feet coordinating* 
 If, however, the starfish is inverted there ia little likelihood that 
 the impulse rLll include coordination of all the tube feat, even after 
 the severest pinching* The only difference between the animal* in these 
 two positions io that the tube feet of the inverted starfish are not 
 executing the step reflex because there is no contact stimulation to set 
 it off* I am inclined to think therefore that a state of orientation 
 spreads much raore rapidly <*here the tube feet are executing the step 
 reflex than -ziiere thoy are not* Thtj|_l_a Iso true of Pisa a tor to a lesser/ 
 extent, but in cane 1 of Pycnopodia it seems to make but little difference 
 whether the tube feat are in contact with the substrate or not* The 
 coordinated impulse is easily initiated and very active in this animal* 
 
 the common or usual manner in which the coordinated impulse 
 is formed in starfish is, X think in general accord with the above out* 
 
 ~VX+:VV >-> \/~ 
 
 line* There *ra very many species of starfish, each differing more or 
 less in its structures and functions from the other so that ideas de- 
 rived from the study of five or six epeoies might not fit the behavior 
 of all of the thousands known to science* 
 
 1 have sean PycnoBo^A* ?jfl-j 9aatg i r ' Aftt^rinA gjid, ^japt^ria.s, 
 regularly orient to**ard or away from contact and charaleal stimulations 
 (mussdl juice or dilute acid) in the manner outlined above, and 
 a bsam of direct sunlight waa thrown, on the eya-spot of 
 the response was analogous to that to contact. 
 
 Orientation .*& result gf. stimulating thji d 
 ox a. general stf^m^l^ti^n of. ajjj^ thja tube 
 

-34- 
 
 The responses of the starfish to light # have been divided 
 by Plessner (1913) into two categories those (both positive and nega- 
 tive) in which the eye spot aots as the receptor and those in which 
 the receptors are distributed over the surface and connected with der- 
 mal nerve net. Inasmuch as it is the whole surface which possesses 
 these receptors and not merely that at the tip of the ray, it would b 
 well here to look into the qualities of the orientation of the tube 
 feet and their coordination that can be brought about through stim- 
 ulating the body wall. 
 
 In starfish which are suspended and the body wall at one 
 side of a ray stimulated by gentle contact I have observed that th 
 tube feet in that region show a tendency to orient themselves in the 
 direction of the stimulus. Upon increasing the strength of the stim- 
 ulation of the body wall, the tube feet near the stimulated area under- 
 go retraction which spreads in proportion to the strngth of the stim- 
 ulus. I have s^en no orientation of the tube feet directly away from 
 the stimulus even though the stimulus be graded in intensity as care- 
 fully as possible. The response is either orientation toward th 
 stimulus or retraction. 
 
 In the above experiment we have an explanation of a positive 
 response to a dermal stimulation* A negative response can be regarded 
 on the above hypothesis as a positive reaction toward the unstimulatad 
 side, if it should indeed prove to be a fact as indicated above that a 
 direct response to dermal stimulation is only positive in its sense. 
 Thus we may suppose that the tube feet are oriented toward the side 
 which receives optimal illumination, rather than that they are oriented 
 
 # The older observers on the responses of starfish to light have 
 divided themselves into two schools* One of these schools regarded. the 
 eye spot as a light receptor and in it may be listed Romanes and Sw$rt 
 (1181), Oraber (1885), Preyer (1886), Bonn (1908). The rnorphologists 
 favored this view also. The second school regarded the light receptors 
 
- 
 
 , 
 
 
 
 I , i 
 
36* 
 
 as in the derails or tube feet. Mangold (1908). Cowles (1911a), Mast 
 (1911), and others adhered to this view more or loss explicitly. The 
 ingenious experiments of Plesner (1913) hare made it seem quite proba- 
 ble that the starfish responds to direct illumination of the dermis and 
 that the eye spot receiver stimulation from distant areas of Hoht or 
 shadow to which tUa starfish res onds also* This results in a very 
 puzzling aggregate of reactions as the controversy attests* 
 
 away from the side that is in a state of sub or super optimal illumina- 
 tion, 
 
 Significance of the negative b r ehavipr ojf the, jj-golatod ray u , 
 
 The negative behavior of the isolated ray, is, as has been 
 
 long inown, much less definite than that of the whole animal* Romanes 
 
 and Kwart ( 1831, p. 1356) state that "Single rays detached from the 
 
 organism crawl" sometimes a^ay from injuries, but they do not invar- 
 
 A 
 
 iably or ovan generally seek to escape from the latter as is so certain 
 to be the ciae with the entire animals"* In confirming this it was 
 found that a migrating ray which had been isolated, wffttld give Very 
 irregular responses to stimuli which would cause negative behavior in a 
 normal animal* A negative response to pinching or prodding is the ex- 
 ception, -rather than the rule in the behavior of isolated rays* This la 
 to be exp30t3d in the light of what h*s been said about the nature of 
 the negative response because the "rays opposite the stimulus" are not 
 there to unfailingly initiate/ a migration away from the stimulus* 
 
 _ . ^ 
 
 BSHAVIQR pv TjB STARFISH 1H5W UKDSH Tflft ISFLUI2TCS 0? 
 
 IMPULSS 
 
 Having studied the factors which govern the formation of the 
 "unified" impulse we shall now turn our attention to the behavior of an 
 animal under the influence of this physiological state, first taking 
 up the factors which cause a change in the "physiological anterior" 
 and factors which cause a change in the direction of locomotion of the 
 starfish by a rotation of the body as a whole without changing the 
 anterior rays. 
 
 the factors which cause a change in the physiological anterior 
 

 
 
 
 t mo 
 
 ' MM 
 
 i*:. 
 
 
 
 fit 
 
 
 EYO^Dfft 
 
 #* 
 
 T 
 
-36- 
 
 are essentially the same as those which determine the anterior 
 as the impulse is being foriae* and operate through the 
 same mechanism, .vith respect to the sense of the reaction which thy 
 elicit thay can therefore be grouped into (1) the positive and 
 $2) the negative. With respect to the receptors on whicn they 
 perate they can be grouped into (1) those acting a the dermis and 
 directly on the tube feet and(*)those acting on the terminal tube 
 feet of the rays (or eye spot which is a modified tube foot). 
 Such Common factors in the environment of the starfish Contact 
 chemical stimulation and g light have been seen to affect the 
 Unified impulse in the uncoordinated starfish in one or more 
 of the above mentioned ways and it will be seen from the following 
 
 that they affect the coordinated impulse once it is started in 
 
 ' 
 the same sense and in the same way* 
 
 Po si tive reaction to r contact 
 
 tOhen. 
 
 $ one of the ray tips of starfish migrating actively 
 under the influence of the "unified impulse" bruslWagainst the 
 side of the aquarium the tube feet at the end of this ray luve 
 been seen *p stretch out actively, those behind them coordinated and 
 soon the direction cf locomotion changed and the animal was 
 walking up the side of the aquarium 
 
 Iterative reaction to_ con tact 
 
 On pinching one of the rays of such a,locomotor starfish, 
 serial retraction or inaotivation of the tube feet will ensue 
 spreading more or less among the tube feet, but last and least 
 effectively to the tube feet of the opposite side of the starfish. 
 The lalle-r y-^-wie <a4\ViTy first and orient mor 
 
 nearly in the direction of the ray on which they are borne i.e. 
 away from the source of stimulation. The tube feet behind these 
 coordinate themselves with them in the same direction so that the 
 coordinated impulse (to ;# away from the stimulus ) spreads 
 

 t M' 
 
-37- 
 
 baok about as quickly as the tube feet become active again. 
 
 Chemioal stimuli # and light (acting on the eye spot) have 
 
 also been seen to affect the loooraotor starfish in a way wholly anal- 
 
 # Romanes 1883 states that all of the under side of the star- 
 fish is sensitive to odor (chemical stimulation) while Pro.uho (1890) 
 localized these receptors in the terminal tube feet of the rays. 
 
 ogous to the above. 
 
 Physiological a^s distinguished from physical orientation. 
 
 I have described above such changes in the direction of a 
 looomotor starfish as involve also changes in the leading ray,- that 
 is the animal may be going in the direction of a certain ray before 
 the change and ia the direction of the opposite rays after change. It 
 is a matter of common observation, however, that crawling starfish 
 sometimes change their orientation by a rotation of the body as a 
 whole without changing the anterior ray. This is a less common method 
 of changing direction, and is said (Bohn 1908) to-be more frequent 
 anng large and stiff specimens than among small active ones. 
 
 Orientation of this kind may be called "physical orienta- 
 tion" to distinguish it from "physiological orientation" which involves 
 a change of the leading ray. 
 
 Physical orientation may involve three f ictore, any one of 
 which may be more or less completely predominant. These are:: (1) 
 Direct orientation of the leading ray or rays to one side: (2) ao 
 celeration of the tube feet of one side of the starfish and a conse- 
 quent swinging of the anterior rays in the opposite direction: (3) 
 the retardation of the tube feet on one side of the starfish and the 
 consequent swinging of the anterior rays toward the same side. 
 

 $r 
 
 
38 
 
 Direct orientation of the leading ray or rays to one side 
 is dependent upon a unilateral stimulation of either the dermis, 
 the eye spot or the tube feet of these rays and a consequent 
 orientation of these rays toward (or away from?) the stimulus. 
 If the stimulus acts also on the rays that are situated on the 
 
 side of the starfish from which the stimulus comes, the anterior 
 
 ^ X 
 is apt to be shifted (Pless*er 1913) to these arms b^t if it 
 
 acts only on the side of the anterior arms it is more likely to causa 
 a rotation of the animal as a whole. This is dependent upon 
 the angle of the stimulus to the direction of the starfish and 
 various other factors that have been analyzed by Bohn (1903). 
 The relative acceleration and retardation of the lateral 
 arms is of course a necessary result of the above described 
 lateral movements of the anterior rays. As a result of 
 stimulation the same factors which we have discussed above act- 
 ing in a positive direction on the tube feet, dermis or eye 
 pot would cause acceleration and in a negative direction would 
 cause retardation, provided the stimulus did not reach the 
 more sensitive (to a direct stimulation) tips of the anterior 
 or posterior rays. A mechanical obstacle to the progress of the 
 rays on one side of the animal will result in a change in 
 orientation that may or may not involve a change in the physio- 
 logical anterior. This, however, will be taken up in connection 
 with the"dviation reaction" and the breaking up of the 
 functional unity of the coordinated impulse. 
 
 GJiNflRAk CONSIDERATION COORDINATION 
 
 t 
 
 The categories into which we have analyzed the reactions 
 of the locomotor starfish are not the separate and distinct 
 unities that they appear above. All of the factors that we 
 have recognized are usually at work at one and the same time* 
 

 
39- 
 
 They are nicely balanced against each other and any stimulus which 
 upsets the balance by adding to the strength of one factor or taking 
 from the strength of another factor results in a more or less radi- 
 cal change in the behavior of tne animal. It is often difficult, 
 moreover, to discern the cause of a change in behavior, eo delicate 
 is the balance between the different factors, and so impossible is 
 it to keep track of the changes of fatigue, hunger, etc*, that play 
 an important part in the relative irritability of the animal as a 
 whole, and of its different parts from time to ti;ae. An analysis 
 of the behavior of starfishes, based upon observations and experi- 
 ments on only four or five species, can not pretend to completeness 
 or to a generality covering the whole group of Aateroidea* (See 
 kangold 1908 on the self burying reaction of Astropefltfpl* 
 Theories ojf the moohaniap gjf coordination.* 
 It is probably true that all starfish locomotion involve! 
 in some of its phases at least a "unified impulse" among the tube 
 feat in various p-arts of the body* 
 
 The mechanism of such coordination is of course very com* 
 p lex. As cording to Von Uexkull, vn the sea urchin it involves the 
 functioning of many nerve nets, connecting and supplying with simi- 
 lar "quantities" of "tonus" homologous parts of the various coordi- 
 nating organs (tube feet, spines etc.,). Pending adequate histolog- 
 ical investigations it would be well to state as an hypothesis that 
 since homologous parts of coordinated tube feet act in almost ex- 
 actly the same manner they are probably connected by nervous paths 
 of lower threshold than are non homolgous parts* 'Hie value of such 
 speculation, however, is dubious, and it is better to keep within thi 
 data of physiology in evaluating the coordinated impulse, since the 
 morphological data is wanting* 
 

 
 
 
 
 
-40 
 
 Orientation of re t rao ted tube feet and the independence 
 mechanisms of orientation and that o withdrawal ojr stepping* 
 
 It has been shown (dole 1913} that the coordinated im- 
 pulse may retain its orientation even after the starfish is removed 
 from water and held inverted for two minutes* This procedure causes 
 the retraction of the tube feet (in Pisaster) and the droping of 
 the arms aborally* vhen put back in the dish of sea water, the 
 animal usually walks in nearly the same direction as before, Thii 
 persistence of direction and the fact that the tube feot are quite 
 retracted after each step, indicates that the mechanism of retrac- 
 tion and extension, of which as we h*ve seen, the step reflex is ft 
 modification, is, perhaps, in no way dependent upon or implicated 
 in the mechanism of orientation* The only point of contact of these 
 two mechanisms is the fact that they both act upon the tube foot* 
 In the locomotor state then oveiy tube foot is oriented, whether it 
 
 *t 
 
 be retracted or not, but retracting and extending in such tube feet 
 are accomplished usually as parts of the step reflex* 
 
 B3HAKING UP TMS SoORPlffATidD BiPUL^ INTft 
 AR^AS IN WHICH ms TUM RCST A!f 
 IN 
 
 Perhaps the most puzzling thing about the unified impulse 
 is the fact that under certain conditions it may be broken up so 
 that it may exist in only a part of the starfish, or tube feet of 
 different parts of the animal became orientsd in different direc- 
 tions* 
 
 Adap j tj.T; j ene j a. j s 
 
 In case of some types (Jennings 1907) of the righting 
 reaction, and in going around an obstacle this orienting of 
 

 
 
 
 
 
 
 
 
 r 
 
 
 
 
41- 
 
 the tube faet in different parts of the starfish in different 
 and sometiijios oppostie directions is highly adaptive in tint it 
 is the only tray the aot could be accomplished* 
 
 Thus in the above diagram, fig* 14 w*iich illustrates a 
 frequently observed type of righting jpeaotion the rays labeled a, & 
 have doubled under and ire migrating in th<j direction of the arrow* 
 The raya labeled o. e, under the influence of the same unified Impulaa 
 have turned in the same dir action but migrate, aftsr having turned, in 
 the opposite direction, thus crossing over the am* a. lj> and complet- 
 ing the somersault* As soon ms the righting is complete the rays 
 o. 2 * th ABM> direotion as the rays a. b. 
 
 
 t~ 
 
 the swirfi^ in position 1 it *as mov- 
 ing in tha direction of thT arTO 1 * and all of the tube feet were 
 oriented in this direction. However, when coming up against the 
 obstacle (3) the tube feet of each ray immediately changed their 
 orientation to the direction indicated by the arrows at the tips of 
 the respective rays. This results in the 
 
 

 
 
 
 hI 
 
 ^ 
 
 
 
 
-42- 
 
 animal neatly avoiding the obstacle and migrating off in the 
 direction indicated by the upper arrow. This is a very interest- 
 ing reaction and has been made the subject of careful study below 
 in an effort to discover the factors concerned in this breaking up 
 of the coordinated impulse* 
 
 Mangold (1908) has described an observation in the slen- 
 der armed Luidia ciliaris in which the animal was seen to have an 
 arm bent so that coordinated tube feet, all extending in the same 
 direction, were some extended out to the right of the ray, some 
 parallel wi th the ray and some to the left of the ray. 
 
 If we are to explain this very puzzling behavior from a 
 physiological standpoint we can not merely point out its adaptive 
 or regulatory value, we must attempt an analysis of iti mechanism. 
 It is futile also, to conjure up a complex "center" in the nervous 
 system which acts as eoordinating mechanism or presiding regulator, 
 orienting the tube feet of various parts of the body in such a man- 
 ner as to best accomplish the act of the moment. Ste iner (18 98^ 
 hypothesizes a "righting center" and Preyer (1886) "centers" for 
 various activities* There is no structural basis for such an assump- 
 tion # and it is not in accord with observations on the behavior of 
 
 # Spix, (1809) described a nervous system for the starfish that 
 would satisfy such an assumption. Unfortunately, ,however, it proved 
 to be the system of gastric and hepatic mesenteris filaments. 
 
 Acceding to Baudelot (1872) who gives an historical resume of 
 the earlier morphological literature the subject became so oontrover- 
 sal that A. H. Quatrefages (1842) made the statement freely trans- 
 la tad as follows. "Naturalists of great merit have come to such 
 diverse conclusions as to the significance of the various systems 
 of (Bchinoderm) organs described as nervous in function that I have 
 decided to remain in this regard in a state of philosophical doubt." 
 

 
 v II 
 
 
 
 
 
 aiej'i 
 
 901 
 
 
 
 -rsrifc 
 ) lo 
 
-43- 
 
 the tube feet which seem to indicate that they all act very much 
 like their neighbors, but with too much independence to lead to the 
 belief that they are subject to the control of a higher center. Tube 
 feet act only in response to stimuli which affect them or spread to 
 them from neighboring tube feet, 
 
 Possible physiological explanation in the traction on the 
 tube feet resulting from the movement of the rays over the substrate, 
 
 It seams to me that the only constant factor that could ac- 
 count for the behavior observed, is the traction of the substrate on 
 the tube feet. This traction is the mechanical result of the move- 
 ment of the starfish over the substrate, (See Cole 1913lr), 
 
 Thu^B Mangold's starfish (fig, ) is moving in the direction 
 
 of the ant>w. The various tube feet may receive stimuli from the aub- 
 
 I ***i 
 
 vx <y 
 
 strate which result in their orienting this direction. 
 
 Similarly the righting starfish has set in action by the 
 activity * the rays a and b_ (fig. 14) a somersaulting motion on a 
 horizontal axis. This results in pulling the rays, c and e in the 
 direction of the arrow tnat indicates their motion. It is this trac- 
 tion that may orient the tube feet. In this connection it is to be 
 noted that if the rays o and e do not droop down to the substrate but 
 are carried over at a level of or above the disk ( as is more often 
 the case) their coordinated impulse does not reverse but remains, as 
 indicated by the parallel extension of the tub feet, in haroiony with 
 that of the rest of the animal. 
 
 In the case of the deviating starfish, the axis of the rota- 
 tion that is involbed in the avoiding erf the obstacle is of course the 
 obstacle itself. There is, in th progress of the reaction first a 
 pushing against the obstacle which involves cessation of locomotion on 
 the part of the rays on one side of the body, but its continuation 
 ( or quick resumption after temporary cessation) 
 

 
 
 
44 * 
 
 on the other, perhaps the stronger, aide. A this continues, 
 due to th comparative rigidity of the animal, there 10 a pull 
 
 (h 1 ^ 
 
 in the direction of the arrows (at the tips of the rays) to which 
 pull the tube feet seem to coordinate themselves. 
 
 Direct pull, exerted throu^i the substrate by the movement 
 of the animal and acting on the tube feet, can,assuming that 
 it orients then, account for the above described behavior. We 
 nail now turn to the evidence for and against the contention 
 that the pull of the substrate does orient the tube feet* 
 
 Direct evidence invoonsluBive. 
 
 The obvious way of testing this is to slowlynpull the animals 
 over the stubs tra to (see Cole 1913&) and ascertain whether a 
 tendency to locomotion in this direction could be built up* 
 About forty treats were made with rigid non-looomotor animals* 
 The tube feet at first caught hold and clung to the substrate* 
 This became less and less manifest and the rigidity of the 
 myodermal sheath gave place to the flexibility that usually 
 
 M* 
 
 accompanies locomotion* Locomotion followed, however, less than 
 
 s^> 
 
 half the trials, the animal more often settling down obstinately 
 
 Tt> W H/cf} 
 
 in the place it was pushed t*. 
 
 o 
 
 Shen the locomotion did fallow it was, unfortunately, in 
 every case but one in the QPPO gi te direction to the puXl. It 
 continued for a few cm* only, <ahen the animal would settle down into 
 
 the rigid state. The one animal that crawled in the direction 
 
 it 
 fe* was pulled, continued to crawl all day* 
 
 These results were complicated by the effects of contact 
 stimulation of the dorsal surfaces which induces close attachment 
 and cessation of locomotion. Hie reactions of the animals, then 
 for the most part may be considered a result of this stimulation 
 rather than a result of the pull. 
 
. 
 
 
 
 
 . 
 
-45- 
 
 X have in fact been unable to manipulate the starfish so 
 as to exert a steady pull in any one direction for any length of time 
 without causing the tube feet to attach and hold on, a tendency -which 
 then spread to other tube feat and inhibited any coordinated impulse 
 that mi/7;ht havo resulted. Later, moraover, on certain occasions they 
 have been observed to retract and be entirely inactive, 
 
 I h-jve manipulated the animals by slowly moving the sub- 
 strates on ^hioh ona or two rays were ?ra Iking and have manipulated 
 tha.-i by means of nourotoiaized or anaesthetized rays but h^ve not been 
 able to do 30 T vi th enough dalicaoy to avoid stimulating the tube feet 
 to become attached or completely retraotad, I ara inolinad, therefore, 
 to consider those results irrelevant rather than evidence against the 
 possibility thit the substrate may have an orienting influence upon 
 
 the tube feet. 
 
 W**t \?'i 
 
 IS vi dance frog neurotomizad animals. 
 
 I f the substrate can orient the tube feet by exerting a 
 directive puXl on them through the movement! of the animal, ^e might 
 expect to find that if one of the posterior arms of a loconotor star* 
 fish -?ere neurotoiaized, tnere might be coordination brough about by 
 the factor in question* Several experiments were performed with it 
 in view to teat this hypothesis, the results of which were complicated 
 by the marked tendency in the injured animals to attach closely and 
 firmly to the substrate. 
 
 The operation was performed on a large, active Pyonopodia* 
 At first the tube feet on the injured ana attached but the movement 
 of the animal wrenched the tube feet loose leaving in one or two caae 
 the dis." affixed to the substrate* As the locomotion continued the 
 tube I3et stuck less and less 
 

 
 
 
) I 
 
 tightly, until they behaved very much like they do in 
 ordinary but rather inictive locomotion. The arm being very 
 flexible, coordination did not occur when the neurotoraized 
 arem *as anterior, because it bent around and under before the 
 tube feet let loose. Some throe or four tours ; fter the 
 operation the tube feet in the nvuromo tized arm were all 
 retracted and the arm practically motionless. A week later the 
 wound seemed to have healed and the arm to hare regained its 
 natural movements. 
 
 When this experiment was repeated on Pisaater. the animal 
 remained stationary for fiw minutes, the neurotoraized ray, 
 affixing itself rather firmly to the substrate* A* the end 
 of this time the other rays were seen pulling in the direction 
 of their former anterior, away from the neurotomized ray. Some 
 refractory tube feet were seen attaching to the substrate, which / 
 were wrenched off by the activity of the uninjured arms. One 
 left Jte disc behind. Refractory feet became fewer and loss 
 refractory. In one minute coordination was complete, though not 
 very active. The animal walked quite rapidly the length of the 
 aquarium. Locomotion seemed normal except that the nsurotomized 
 arm was contracted and rigid. It was always behind or obliquely 
 behind in locomotion* 
 
 It might seem possible therefore that coordination of the 
 tube feet is not wholly dependent upon the presence of an insect 
 nervous system* If such stimuli as cause the attaching reflex, are 
 carefully excluded coordination may be established, across a -ut 
 nerve cord by the traction of the other anas. 
 
 When the neurotomized starfish had tone to rest it was 
 observed that the four intact rays were stationary while the neuro- 
 moized ray walked xautit about in the sector between the adjacent 
 stationary rays. I Eion prodded the starfish and threw it into a 
 

 
 
 as 
 
-47 
 
 vary intenaly appreaeed atato. The neuro tomized ray continued 
 a a before actively moving in its own sector. The gilla were 
 retraoted and tho pedioellariae open, OTer the whole atarfiah 
 while in the region of the out and beyond the gills were out 
 normally and the pedioellanae at reat. On prodding the neurotomiaed 
 ant the gilla drew in, the pedioellanae stood out and opened 
 and the tube feet held fast. This last reaction paaaed off 
 and the noufcotomized arm started locomotion again in ita aeotor. 
 The gilla and pedioellaria remained in the irritated state so 
 
 that the out did not deraark two different areas of gilla and 
 
 and 
 pedioellaria aa it had before/<*s it did now with the tube feet* 
 
 X believe, therefore, that neural oonneotion for the 
 apreading of an Impulse aoroaa the out, either through the 
 dermal nerve net or through an unout portion of the ambulaoral 
 oord, was entirely absent. 
 
 The eaaentials of these eaperiments were repeated on a 
 nuaber of animals, with very similar resulta. Asterina 
 
 responds in thia way but rather less completely than Pisa a tar. 
 
 \r\ w f?o~ 
 An active atarfiah with a e anterior (aee pr"T*5) 
 
 wma picked up quicly and the raya bod neurotomized* The 
 animal was aet on the aide of the aquarium with the intact raya 
 (a ) directed downwards. Locomotion followed a e down the aide 
 and across the aquarium. S e d presented refractory tube feet 
 and locomotion was jerky as these tube feet were pulled loose. 
 Later, when the animal had progressed about 6 cm coordination waa 
 fairly well established but not very active. Aa the refractory 
 tube feet were pulled loose they retracted and did not react at 
 ally for sometime. Neighboring tube feet, however, showed 
 diminished tendency to attach tightly and were more apt to 
 coordinate. Locomotion waa alow at first but later more rapid. 
 

 
 
 
 
 
 
 
 
 
 
 i ; ** h 
 
 
 
 f.' 
 
 
 
-48- 
 
 (-near the taO 
 
 X next neuro tomi xed ^ea oh arm of a rather large starfish that 
 was not very active. X "started" it on the side of the aquarium 
 vath ita former "anterior" downward. Locomotion continued down 
 the aide until the disc was about at the angle of the wall with 
 the floor of the aquarium* At this point, the animal assumed the 
 rigid state and would orawl no farther* 
 
 This experiment was repeated on a amaller and more active, 
 specimen. Locomotion down the side was more active, the 
 
 (former anterior) arm* taking up the locomotion quickly and by 
 pulling, in harmony with the force of gravity, forced aoertain 
 amount of coordination in the other rays* There were a few 
 refractory tube feet in each of the rays, each ray shwwing a 
 tendency to migrate toward its own tip* 3hen the animal reached 
 the angle of the side with the floor of the aquarium the locomo 
 tor impuloe was so well established that crawling continued 
 across the floor of the aquarium and up the other side* If 
 an obstacle such as my finger was placed between the two anterior 
 rays and held stationary, two responses were observed* In two 
 oases a normal deviation reaction ensued, but the more frequent 
 result was a stoppage of locomotion followed after a variable 
 length of time b a resumption of locomotion in some other 
 direction* 
 
 The starfish wa* then taken up and stimulated harshly on 
 the various rays* The animal assumed tho rigid state -then set 
 down the tube feet being tightly attached, and remained in this 
 
 state for some time* The rays, that became active first were not 
 contiguous, X -a4-o, while b d and e remained attached* A and o^ 
 moved a out in their sectors at random all the afternoon* The 
 next raoraing the starfish was in a mo rib und condition but 
 had migrated across the aquarium during the night* 
 
 The essentials of these experiments were repeated many 
 

 
 .- 
 
 - 
 
 
..-49- 
 
 times with results thatvaried between the two examples cited* 
 It was found that if the manipulation was rough or unnecessarily 
 prolonged, the animals would become rigidly attached and would 
 not locomote for some time or at all while some animals refused 
 to coordinate with even the gentlest manipulation*^ 
 
 # Opinion on the necessity of an intact nervous system for 
 eohinoderm coordination seems divided. Romanes and Jfiwart (1881) 
 and Cole (1913) record some slow coordination between parts 
 on opposite sides of a cut in the nervous system, while Husso 
 (1913) believes that coordination may be absolutely normal with 
 the oral nerve ring removed, Tlark (1890) states that the 
 movements of the tentacles in Synapta and coordinated movements 
 of the body muscles are not destroyed by cutting the nerve ring* 
 See also Grave jqoo on Qphuira breviepina 
 
 Among those who report the opposite results are Vulpian 
 (1862) ,:rukeuberg (1881) DeMoor an* Ohapeaux 1891 Loeb (1900) 
 Mango lr (Wow i4W ) fJtoore (19100, 1910te) ete 
 
 Trom these experiments, and those on trie righting of 
 neurotomixed animals which will be described later, I think that 
 it can be safaly concluded that while there is no neural or 
 neuroid" (Parker 191$} transmission pastja cut in the am ulacral 
 nervous system, there may be a certain limited amount of coor- 
 dination between parts separated by such a cut brought about 
 -mwfafiHe*** through there mutual relationships to ehe substrate* 
 
 yridenoo froyi the behavior oj[ ffie animal when its, part? 
 aye placed on. aet^rftte, p n ubst rates,.' 
 
 We shall turn now to such indirect evidence as bears upon 
 this point from the behavior of an animal on separate substrates 
 and A a quantitative analysis of the mechanics of the deviation 
 reaction. 'Uiese methods, though indirect do not cause the 
 attaching reaction* 
 
 The rays of an active starfish that is not in the coordinated 
 
 as 
 state, /has been seen above will migrate toward their tips, into 
 
 free floating glass tubes* If however before suspending and before 
 the floats are presented to the ry. * anim al * in * teu 
 
50 
 
 Of active locomotion, the rays that were anterior will crawl 
 on into the tubes while the ray that were posterior will 
 start to crawl out of them* Usually before one of these rays 
 f leaves go its hold on the float, or at any rate soon afterward, 
 the impulse in thia ray is reversed and it la aeen to be 
 active in its migration toward its own tip, regardless of the 
 direction in which the other rays are crawling. If now the 
 tubes are removed from their floats and set on the bottom of the 
 aquarium, -vith the tip of a ray in each, the coordimt i impulse 
 is quickly re -established and the animal migrates back ad forth 
 within the confines set up by the ends of the tubes* After 
 extensive experimentation with the reactions of Pisaster in 
 these floats, I have very seldom seen the unified impulse appear 
 when the floats were free to move separately, and having 
 appeared it seldom lasts more than a minute or two* It appears 
 quite promptly and lasts for a long time ( an hour or more) 
 if the tubes are not separately moveable but are resting on the 
 bottom of the aquarium* 
 
 Supplementary experiments were carried on with flat free 
 swinging substrates* One, two or three of the rays were put on 
 the substrate and the others allowed to han^ over the iide on 
 the floor of the aquarium half a cm* below* The part, on one 
 substrate was often seen to migrate while that on the other 
 remained stationary, and they were not infrequently seen to 
 migrate in different directions. Of course this would not be 
 likely to happen if the substrates were not separately moveable, 
 
 From the above experiments it would seem that a factor in 
 the unity of the coordinated impulse is the unity of the 
 substrate or rather of the animals relation to the 
 

 / 
 
 
 
 If 
 
 
-51- 
 
 # One might state the oase rather paradoxically in metaphysical terms 
 by saying that the animal* 8 soul or entelechy, or some part of it at 
 least resides in its substrate. (See Dreisch (1908) Steme (1891).) 
 
 Therefore, if the activity of the animal caused the substrate to move in 
 one direction with reference to one part and in another direction with 
 reference to another part, as is the oase in the righting and deviation 
 reaction, we might expect that the unified impulse would be broken up in 
 certain detaitninate way*. 
 
 Deviation reaction not interfered with by_ cutting ne rvous con- 
 nections with interradial a rea 
 
 That the coordinated impulse is thus broken up by mechanical 
 traction in the deviation reaction, is made likely by the fact that 
 the reaction is perfectly normal even after the nerve net on the out- 
 aide of the epidermis was out through between the obstacle and the am- 
 bulaoral nervous system. This, of course prevented any stimulus from 
 the contact of the starfish with the obstacle reaching the tube feet, 
 but did not affect the mechanical factors in the relation of the sub- 
 strate with the tube feet. It is therefore to be concluded that these 
 mechanical factors play an important role in the deviation reaction. 
 
 Deviation reaction not elicited by. prodding interradial area. 
 Moreover, if the nerve net between the bases of the two an- 
 terior rays be stimulated by jabbing it quickly with a knife or a blunt 
 instrument, the deviation reaction w.ill not follow* The aniaal will 
 either continue undisturbed, stop and then continue or go into the at- 
 tached condition and remain so more or less permanently. The first 
 response is by far the most common if the specimen is normally active 
 and not stimulated too harshly. I have never observed a marked change of 
 direction as is seen in the deviation reaction to say noihing of the 
 
- 
 
 - JSi 
 
 
 
 '. Oil! 
 
 . 
 
-52- 
 
 ooaiplioited coordination cf movements thit are involved in the 
 deviation reaction* 
 
 iMMl f ^ tn *^' y< * aspects pf the "deviation push" oji different 
 substrates and witft different feints on the animal vary *ith 
 mechanical conditions while quantitative aspects olT oontaot 
 stimuli required to initiate the negative reaction dft not. 
 
 It was thought that the auount of push which the deviating 
 animal exerted upon the obstacle -whenjoonsidered in oonneotion 
 with itB pulling Ability, and other reactions might thro light 
 upon the mechanisms of the deviation reaction. The amount 
 of push was measured by attaching the obstacle, a lever, swinging 
 freely frofr a rigid fulcrum by a thread to the recording 
 spring above described* The push, then was recorded as the 
 height of the curve, written on the slowly revolving drum* The 
 appearance of the curve was as below for the different species 
 studied* 
 
 The push continues to increase until t&a deviation begins, that 
 is, until the effectors (tube feet or spines) on one side of the 
 body begin to reverse themselves and the rotation around the 
 obstacle as an axis is initiated* From then on there is an 
 irregular decline in the push until the naoDdb animal is free 
 of the obstacle* ith the drum running at the same speed, the 
 shape of the curve as well ae its height is dependent upon the 
 aotivity of the specimen studied* This was taken into account 
 so as to get results as comparable as possible* If The sea urchin 
 
 1 1 roa/ry l o c en trp t ep franoisoanua was found to be mowing on sand 
 by means of its spines only. In deviating around an obstacle 
 
ttft ol 
 
 
 I fifti 
 
 
 i 
 
 14, 
 
 . 
 
 
 
 . 
 
 
 ^-i*S 
 
 ' 
 
 "^ .: 
 
 ll le MUHHI 
 
-53- 
 
 it takes the same course aa a starfish. This ease is oited since the 
 spines of the sea urchin do not attach and their behavior in this con- 
 nection indio.itss a rathar striking similarity between the physiology 
 of the spine and that of the tube foot* 
 
 The value of the "deviation push* of this specimen, was found 
 to average 15 g. This was increased to 17 g when a load (about 40 g) 
 was placed on the dorsal side of the animal. The "pulling ability" was 
 found to be (average of 6 trials) 10 g unloaded and 15 g loaded. Allow- 
 ing for a certain amount of fatigue in the later trials the "pulling 
 ability" was found to be approximately equal to the "deviation push". 
 
 The same relationship seams to nold with Pyonopodia. As 
 seen above the pulling ability averages 47 g. the deviation reaction 
 (average of four trials 60 g 45 g 60 g 30 g) is 48g. These are in- 
 creased to V2 and 105 g respectively by loaning the animal with 80 g. 
 of glass-flare* If the animal is placed on sand the values are similarly 
 related to each otli.sr but are reduced as follows* Pulling 15 fim, pull- 
 loaaed -with 80 gnu 32 gm, deviation 29 gm, deviation loaded 35 cm. 
 
 Due to the fact that Pisa a tor and Asterjna are able to pull 
 very much harder in proportion to their size than are the sea urchin 
 or Pyonopodia and since this pull is due to the constant increase of the 
 attaching tendency correlated with the pull, we find that the deviation 
 push correlates more closely with the pulling ability on sand, taking 
 into account of course its lesser frictional coefficient, than with the 
 pulling reaction on a solid substrate* The average deviation push of 
 Pisaster (about 15 cm in diameter) is 20 g. on a solid substrate and 6 g 
 on sanJ. Agtarina (8 cm) on a solid substrate exerts a deviation push of 
 4 g t but with 4 g. weight on ita back this is increased to 6 g. This is 
 ~*ith the pulling ability of a larger specimen on sand of 7.5g. 
 

 
 - 
 
 
-54 
 
 and on sand weighted (40g) of 15 g (See p. 21). The above study 
 of the mechanics of the deviation does not pretsnt to be statist!- 
 oally comprehensive. The objact is merely to point out that the 
 "deviation push" can be always increased bu weighting down the animal 
 and that in the sea urchin, which uses its spines, and in Pyonopodia 
 which does not attach tightly while pulling hard (Soe p. 22) the pull 
 oan also be increased by weighting down the animal. The relation* 
 ships of pull, and deviation push in the loaded and unloaded Asterijia 
 and i'isastar, are consistent with the above and comparable, quanti- 
 tatively to the pulling ability of the animals, both loaded and un- 
 loaded on sand* 
 
 Thus, the attaching reflex that strangthens with the resis- 
 tance to the ordinary step (see p. 19) does not appear comparably 
 in the deviation reaction. This it seams to me is because the tube 
 feat on one side of the obstacle overbalance in their traction those 
 on the other side, cause a rotation of the animal in that direction 
 and the various tube feet coordinate in the direction of this rota* 
 tion. There is than no resistance to the step bi'.t merely a devia* 
 tion of it in one direction or the other brought about by its relation 
 to the substrata* 
 
 Another fict pointing to the conclusion that the factor* 
 of the deviation reaction have to do with the mechanical relation* 
 ship of animal to substrate rather than with reflexes having their 
 receptors at the point of contact is that if the tips (Asterina) 
 of the rays instead of the dermi. between the rays com* in contact 
 Tith one obstacle connected with the spring recorder the amount of 
 pressure that it takes to cause a change in direction, does not 
 vary if weight is put on the back of trie starfish, Xhe value it 
 about 2.5 g in each case* This shows 
 

 
 
 
-55- ."' , 5i 
 
 as might be expected from the configuration of the nervous system, that 
 the mechanism of the deviation reaction is altogether different from 
 the mechanism involved in a change of direction rhen the tips of the 
 rays are stimulated. In the one case we are dealing with the relatively 
 constant threshold of the receptors in the end of the ray while in the 
 other oase we are dealing with factors that vary with the mechanical 
 data of load and friction. 
 
 In order that the obstacle may be left behind in the devia- 
 tion reaction there is usually a turn of at least 70 which is often 
 recovered from, by the operation of a tendency, whose mechanism I have 
 not worked out, to continue crawling in the saue direction as before 
 the disturbance, even if the action involve an actual change of direc- 
 tion, back, from one assumed as the result of the disturbance* This 
 tendency will also be noticed in connection with the righting reaction 
 (P. 75). 
 
 COORDINATION OF MOV3lteNT8 OF THiS TUBfl F&ST WITH TKOSB OF 
 
 THS ABM AS A WHO 18. 
 
 Il.lus t ra t long o f the tendency of an arm to set itself more 
 ajt right angles tp_ its actively .en-ten teat tube feet, when such move- 
 ments involve dorsal and vent ral fleion and lateral twisting. 
 
 If an active starfish be suspended and a solid object be 
 brou^it in contact with the tip of one of the rays, there will be a 
 movement of the tube feet in the direction 
 

 
 
58- 
 
 
 of the object, an activation of their coordination toward the 
 tip of the ray* This will bo followed, almost immediately by 
 a dorso -flexion of the tip of the ray* She ray oan be said 
 
 ywm&Ai 
 to set ifaalf more nearly at ri^ht angles to the extended active 
 
 tube feet* This reaction has been observed time and again in 
 fisaater oraceuw. Aaterina. pyonopodia. ?,eptaaterias i a. 8 tor bre- 
 vi&inus and -.vaeteriaau As are moot movements 
 
 of the animal it is a product of local reflexes in that it ie not 
 dependent upon connection with the oral nerve ring, but 
 occurs equally veil in active isolated arms. 
 
 . for the gentle contact we aubatitute a harsh tapping of the 
 tip of the ray, the tube faet will retraot and the ray become 
 more rigid and shorter, but without any sign of the dorsal 
 
 flexion* 
 
 w have seen that if a tube foot in the middle of a ray 
 
 be allowed to attach to an objeot and the object be then pulled 
 
 to one side, the tube foot with it, other tube fee* will alee 
 
 owe to the sane side and seemingly reach out for the object 
 
 to which the tube foot is attached* Now if a sufficient 
 
 number of tube feet become oriented in this manner, there will be 
 
 a lateral twist of the ray toward the object* Here again the 
 

 
 
 
59- 
 
 ray oan be said to sat itself more nearly at right angles to the 
 oriented active tube feet by lateral as well as by dorsal movement* 
 
 
 (See also Jennings (1907) description of the taking of food from th 
 pedioellariae by the tube feet)* 
 
 V-" Hie slender armed speoies of starfish 
 
 trosohelii) was suspended and a flat piece of thin celluloid was swung 
 by a thread to the ventral side of one of the rays. The tube feet, 
 oriented rather inactively toward the tip of the ray immediately 
 
 sieze the object and "walk" it in the direction of the base* This 
 
 \ 
 was observed to involve the orientation toward the object of quite 
 
 a number of tube feet both above and below it and the bending of the 
 ray so as to receive the object in a sort of hollow* The tube feet 
 In actual contact with the object are, 
 
 of course, undergoing the step reflex, but above and below, where 
 the tube feet are all directed toward the object, it oan be said, 
 gain that the ray tends to set itself more at right angles to aotive< 
 ly oriented tube fe-st, this time involving both dorsal and ventral 
 flexion* In the region where the tube feet are undergoing the step* 
 reflex, there is no bending of the ray# 
 
 # It has been shown by both Jennings (1907) and Mangold (1908) 
 

 
 ' 
 
 
 
60* 
 
 that as the tuba feet carry a small pieee of food toward the 
 mouth there is a "humping up" of the ray in the region of the 
 food which probably involves the factors desoribed above. The 
 behavior of the tube feet when the animal moves its arm in 
 under the disc as a part of the food taking response (Jennings 
 1907) would be interesting but Z have never been able to 
 induce this response in the species at hand* 
 
 Van^ral flexion pj[ ri.'did.of loured and nicotinlzad starfish/ 
 If a Piaaster in a state of extreme rigidity be inverted 
 there will be as we have seen, a rather inactive extension of the 
 tube feet more or less at right angles to the rays* There will be 
 no orientation of the tube feet at the tip in the direction of 
 the ray* The rays, soon after inverting will lift themselves 
 orally and assume a very symetrioal ventral flexion* This 
 state may continue, in absenoe of disturbing stimulation for as 
 much as twelve hours* If the radial nerves be out or injured 
 near the base* this ventro flexion is apt to be very ouch 
 intensified *o that the steps of the rays come nearly or quite 
 in contact and the animal assumes what Roman** (i3Xt and 
 ftfart (1881) who describe this response aptly call " a tulip lite 
 fora"* This is similar to the state of ventro flexion which 
 Moore (I920a^ # describes as a result of nicotine poioning, and 
 
 / The effect of nicotine on starfish had been desoribed 
 previously by Preyer (1886) and Greenwood (1890) 
 
 wfcioh Z have confirmed for Pisaster, lh chief difference seems 
 to be that the tube feet in the nice tini zed Pfrsastar art completely 
 retracted, while those of the rigid, or of the nourotomized 
 animal show a certain amount of extension but no particular 
 orientation. !Ih strength of the spasm is greater in the nico- 
 tinized animal ffcsw. 
 
 These movements are shown by the isolated ray from both 
 the niootinized (Moore) and the rigid animal* 
 
MM0 
 
 JC jnt *df a* tfvjuru 
 
 .sro/s * 
 
 nB 
 
 . 
 
 HNt* n^VM JB VWR f 
 
 . 
 
 t*(wr- 
 
-61- 
 
 Deaoription of various other oorrelatgd movements o_f the 
 tube feet and 2JS&* 
 
 If an aotive Pisaster be suspended in -water and away from 
 contact stimulation, the rays move about for a while, flexing them- 
 selves doraally and laterally, in m manner that we shall discuss 
 latar, but eventually assume a state of vantro -flexion similar to 
 that assumed by the rigid animal* The aotive animal in vsntro flex* 
 ion differs, however, from the rigid or the niootinized animal in 
 that contact stimulations at onoe set up activities of the tube feat 
 and arms* The tube feat raaot positively to gentle contact stimula- 
 tion and retraot upon sever* stimulation. We have followed the im- 
 mediate responses of the anas to these stimulations, but the positive 
 and negative activities of the tube fset spread to the tube feet of 
 the rest of the animal, as also do the corresponding movements of 
 the arms* 'ihus if the stimulation be quite harsh the tube feet will 
 retraot over the wnola animal and the arms themselves will become 
 shorter and more rigid* 
 
 In oonneotion with the positive response of the tube feet, 
 ir will be remembered that this does not spread as well when the tube 
 feet are free from contaot as it dies when they are executing the 
 step reflex* A weak positive response theo, such as the positive dif- 
 ferential activity of the unstimulated anas in case of a harshly tim- 
 ulatad animal, hardly makes itself notiovable in the suspended animal 
 as it does in the ae&ativt response of the animal locomoting on a sub- 
 strata* 
 
 Ascription oj; th..a formation the coordinated jlmpulse T yhen 
 the tuba feet are free o the substrata* 
 
 A strong positive response, on the other hand, does spread, 
 and in spreading involves movements of the arm, as the following ex- 
 periment will show. An aotive Pieaster suspended and in a state of 
 

 
 
 At 
 
 ? -art* c . t,! 
 
 It 
 
 
 t *- 
 
 ' 
 
 CiJ 
 
 
 .OtfOt^Mf^^ yjj 10 Itf 
 
 * a-."- , .eB'jc^afci avi. . A 
 
/ 16 aide of the aquarium no 
 
 3h^th will. She tube feet at 
 
 tretched out toward the wall, 
 tatfish still farther over 
 V__-- rays a e nexed dorsally and 
 
 began migrating in their own direction. In the meantime the 
 coordination of the tube feot had spread so as to include all 
 the extended tuba feet in animal, which were soon all pointed 
 directly toward the wall of the aquarium* As the tube feet 
 became oriented in this direction, there was a coordinated 
 movement of the rays. Hay twisted to the left and bent over 
 toward the wall, ray & twisted to the right and bent over to 
 the wall and ray g bent directly over the disc toward the 
 wall, isacfli ray was seen to set itself more at right angles to 
 
 Cnin^j^i^d 
 
 the actively -extended tube feet whiOh had become coordinatedly 
 pointing toward the wall* 
 
 As the rays a e continued their activity, the disc was 
 brought closer to the wall and with it, the other arms* As they 
 touted the wall, since the tube feet were oriented in the direo- 
 
 tion of a e beaan executing the step reflex in this direction 
 and the animal started perfectly coordinated and normal locomotion 
 in the direction of ft e (the suspending thread having been out) 
 go rr elation oj these movements with the rj.ffrtinff reaction* 
 The *ove experiment is merely a simplification of the 
 righting reaction of the uncoordinated active Pisaster. If we 
 assume that two adjacent rays initiate the reaction by 
 attaching to the substrate with their ventral sides turned toward 
 each oth^r, the above description will fit the rioting reaction 
 with the change of only a few words. The same dorsal flexion 
 of the initating rays and their migration toward their tips will 
 

 
 
 
 
 4Wf$ 
 
-83- 
 
 be observed. Tna tube feet will alljooordinate pointing in the 
 direction of the initiating rays and' the other ray a will move 
 so as to oome more at right angles to tho direction of the 
 tub* feet. The <aim on the right will twist to the right, Tnd 
 more over in the direction of the initiating raya. The ana 
 on the left will twiat to the left and do the same tiling* ihe 
 arm directly opposite the initiating arms will bend directly 
 over the disk and complete the somersault with locomotion, as 
 ws- shall show later, continuing generally in the direotion of the 
 initiating rays* This as we ahall see is perhape the moat common 
 
 method of righting it tho disposal of the starfish* 
 
 ing 
 Analysis ojf Penning* a Seven types of Hjgh/f reaQi.ion* 
 
 Jennings (1907 pp. 125g ff } however, describes seven 
 sin ty]MM about which the extremely variable righting reaction 
 may be grouped. The first type is: 
 
 l.*The simplest and neatest method is the following* X*o 
 adjacent rays twist their tips in suoh a way that the ventral 
 surface s of the two face each other* 'Ihen the tube feet of 
 these rays attach themselves and throw the starfish over in a 
 neat somersault** 
 
 This is essentially the method described by me above* 
 Jennings description leaves out, hare, the coordinated action 
 ef the unattached arms thou^i he mentions it elsewhere in 
 general terms, and he does not recognize the aprea^ of the 
 coordination among the tube feet nor its relation to the 
 movements of the arms* As above stated this is the commonest 
 method of turning* We shall inquire as to the reason for the 
 turning of the rays toward each otfesr in a majority of oases 
 in connection with our discussion of the righting of the 
 oriented starfish* 
 

-64- 
 
 Jennings sooond type/i* ae follows, 
 
 2. *2he tips of the two adjacent rays may so twist that tha 
 ventral surfaces do not faoe eaoh other, but both faoa in tha 
 same direction. The tube feet then take hold and thro* the 
 starfish over, twisting it about an axis x* whioh passes length- 
 wise through one cf the attached rays* this method of turning is 
 extremely difficult end awkward but is seen at time. Usually 
 however ***> a, third ray takes hold and aids in the turning, 
 the method then forming a transition to that given next** 
 
 I have observed this method of righting only a few times, and 
 variations of it (Type 5 (6) of Jennings) where only one ray 
 attaches a few times also* In eaoh case the coordinated impulse 
 could be seen to spread froxa the initiating ray or rays and 
 involve coordination of the rest of the tube feet and to some 
 extend the arms in the manner described above* She ray that might 
 be expected to attach coordinately (facing) the ray that bends 
 down is usually seen lifted above the substrate and reaching 
 out in the direction of the righting* Locomotion after righting 
 is usually toward the rays that initiate the reaction* 
 
 Jenning*s thr$d type is as follows* 
 3,*?hree adjacent rays attach and remain attached, all 
 pulling throughout the reaction* Usually the animal turns pri- 
 marily by the aid of the t**o outer rgfcs, while the middle one 
 is relatively passive and compelled to double back under as the 
 araimal turns. Often this middle ray walks backward beneath one 
 of the other rays, or the otfcer walks actively over its surface, 
 or there is a combination of these two movements tillthe normal 
 position is reached:, ( A model of the starfish in paper or 
 cloth will make clear the necessity of such movements when 
 three of the rays remain atvtohed^" 
 
 '.mere is no new principal involved here, except that of 
 

 
 
 
 
 
 .c 
 
 , 
 
 
 
66- 
 
 O) 
 
 the passive movement of the middle x ray which will be 
 
 A 
 
 discussed in connection with the fourths type* ifte impulse 
 spreads to the tube feet of the two unattached rays* The 
 coordination of these is followed by their raising up over the 
 diso and moving toward the initiating rays in the Mae way and 
 according to the same principles* as described above* 
 (types 1 and 2), 
 
 The fourth type is as follows* 
 
 4. "Four of the rays take hold, two extending to the right, 
 two to the left. Then the fifth ray, (which we nay eall the 
 posterior one) is lifted straight up and swings directly over 
 till its ventral surface reaches the bottom, while the anterior 
 attached pair walks backward beneath the posterior attached 
 
 pair the latter walking forward over the surface of the (former)" 
 
 f-M if 
 mis type of righting is sketched en p. In oase 
 
 of Plaaster it is more apt to scour if the animal is very much 
 relaxed* The sequence of the events as I have observed it is as 
 follows. The anterior rays twist toward each other and the 
 coordinated impulse spreads over (or is already in) the starfish 
 
 a * in type ! This results in the twisting toward them of the 
 
 p*t 
 lateral rays and the bending up of the posterior ray* fate to 
 
 the relaxed state of the starfish ersoroe other physiological 
 factor which prevents the lateral arms assuming their usual state 
 of ventro flexion, these droop to the atubstrate and become the 
 
 e,e 
 "posterior attached arms" (rays fc in fig] 1 *) How the fadtor 
 
 which causes the moving forward of the back rays xhen the direc* 
 tion of the coordinated impulse* as seen by the activity of the 
 Initiating rays causes locomotion in the opposite direction is 
 the same factor, X think, which amuses the complex coordination 
 of the deviation reaction, I have presented the evidence which 
 leads me to think that the factor in question has to do with 
 

 
 
 
66* 
 
 the relation of the moving parts of the animl to the substrate and 
 a consequent orientation of the tube feet in the direction tf 
 the movement* 
 
 Jennings fifth type is as follows: 
 
 (4) 6* All of the rays attach themselves. How the turning 
 can be accomplished only by the release of certain rays, ihen the 
 Method passes to one of the types already described* 
 
 The method of release as I have observed it is of two 
 kinds* (!) 'i3ie pull of the other parts of the starfish tear loo so ^ 
 attached tube feet* These then retract and other tube feet 
 attach but usually not so tightly as those that were first 
 attached* As this continues the tube feet in the region in 
 question either all become retracted and the ray is pulled free 
 
 (2) 
 
 of the substrate and swung over in the righting, * the tube 
 feet become oriented in the direction of the pull and righting 
 proceeds according to method three or four with possibly a 
 
 (r.e(eA) 
 
 lifting of the looonotor ray free of the substrate* 
 
 A 
 
 Jonninge sixth type has already been described in connection 
 with his second type* 
 
 Jenning's seventh type is as follows: 
 
 *(6) 7 A still more unusual type is seen in the performance 
 of the righting action without attachment of the tube feet of 
 any of the rays* Preyer (1886) and Romanes (1885) have given 
 account of certain ways in which this is sometimes accomplished* 
 The typical tsethod seems for the starfish to raise its disk 
 &igh standing on the tips of all the five rays, then to swing 
 one or more rays over, or one or more under or both until the 
 body topples over ventral side down* In my own observations, 
 the righting without attaching the tube feet was seen only when 
 these were experimentally prevented from taking hold* The 
 starfish then writhed and squirmed irregularly, taking various 
 
trf* flL. 
 
 
 
 
 
 
-67 
 
 bizarre forms, until it had succeeded in getting its ventral 
 side down^when.the squirming cesed, 
 
 The method of righting, described by Romanes and Preyer 
 
 e,c 
 
 seems to be confined to Astropfroien and its allies. I hav 
 never had access to one of these species and therefore shall 
 regard this highly specialised sand burrowing group as outside 
 the scope of the present paper* The peculiarities of their 
 righting reaction are said (Romanes 188$) to be contingent upon 
 the fact that the tube feet are not equipped with suckers and 
 hence do not attach* 
 
 Description the riiditimt reaction as oosurs yhen the 
 tube feet are prevented attaching by inverting the animal o* 
 sand. 
 
 With the animals at my disposal it was thought possible to 
 prevent the attachment of the tube feat*' by investing upon sand. 
 
 2he behavior of a large sluggish Pitas ter .when inverted 
 on earn is interesting in connection with Moors (1916, 1918, 
 1920# 19204) recent observations on strychnine poisoned starfish* 
 The tube feet at the tips of all of the rays of the large sluggish 
 animal* I had under observation extended out toward the tips and 
 the rays bent dorsally, setting themselves more nearly at right 
 angles to the actively extended tube feet* The tube feet 
 however did not attach as they came in contact only with sand* 
 The coordination of tube feet did not spread back very far and 
 the dor so -flexion involved only the distal pats of the rays* 
 Tor some time all five rays remained donro-flexed* 'tfhen the animals 
 we*- placed on 1t$s ventral side on the sand, there was still ?very 
 marked tendency for the rays to 11 bend dorsally at the tips* 
 
 Now hen a similar specimen, large and sluggish, was placed 
 in a dish of strychnine sulfate in sea water 1*10,000 the same 
 picture appeared, with the additional factor t at the tube feet 
 suckers were so paraliaed that they could not attach to a solid 
 substrate* There was then, a tendency toward dorsoflexion at 
 the tip of the rays and a failure of the coordinated impulse to 
 spread readily among the tube feet as a result eiffi.r -f the paralys* 
 of the tube feet by strychnine and of prevention of their attachment 
 on sand* 
 
 These results are probably merely analygous to those of Moore 
 on As ten. an forbeai and tend to demonstrate the many ways in which 
 a givoii response may be brought about in the various pxx represen- 
 tatives of the asteroidea. I have, moreover, so far been unable 
 to get in yisaster the marked dorao-flexion which Moore figures 
 for Aateria lesi. 
 

 WMfiM 
 
 4& qi*y fl'toriftl i ^ypflft^ i^H 
 
 J*LLIL: 
 
 . XJB9C; 
 
 .i^. ".*. aetfr aai41tffTiil vtf'- 
 
 .-. iiii A mftBA:' 
 
 - 
 X^*^nf 
 
 , 
 
 i 
 
 & > 
 fc-x 
 
 
 
-68- 
 
 It would be obviously impossible for the suckers to attach., yet tha 
 animals (Asterina especially) righted themselves quite as neatly as 
 on a solid substrate. Pilaster, however, would not right easily un- 
 less in active locomotion at the time of inversion* 
 
 A specimen actively crawling in the direction of a e 
 (fig. 16) was quickly inverted on sand. The tube feet, which were 
 retracted because the animal was lifted from the substrate, extended 
 at once toward a e. B and moved up orally and twisted toward a e, 
 bent up and over the disk while a s- twisted toward each other and 
 the tube feet, as soon as they came in contact with the sand, began 
 executing the step reflex. Thus each ray moved so as to set itself 
 more nearly at right angles to its actively extended (Oriented) tuba 
 feat. The stepping activity of the tube feet on a a resulted in their 
 doubling back under themselves, so that the tube feet were striking 
 
 <je 
 
 out toward the disk instead of away from it (see rays A fig. 18, 19, 20). 
 The step reflexes of the tube feet in contact with the sand were very 
 active, the ends of the feet plou^iing back through the sand and scat- 
 tering the grains on all sides to a distance of one or two centimeters. 
 The movements thus initiated continued until the rays a e had walked 
 back under the disk and the other rays had moved up over the disk far 
 
 enough to overbalance the animal and complete the somersault. Loco- 
 
 ' ' -* ^^^L^L^L^L^HI 
 
 motion then continued in the direction of a e 
 
 The righting reaction of Asterina on sand is even neata&r 
 than that of Pisastar. This is dua to the very great flexibility of 
 the ray tips and to the strength and size of the large disked tube 
 feet. The animal rights nearly as quickly and easily as on a solid 
 substrata. 
 
 INTJRPRKTATIOH OJ THJ RIGHTING REACTION Aj^ A PHASE 
 
 OF LOCOMOTION 
 3Svide.no a from the mo veme.nt of the tube feet and arms.. 
 
^srlftt** 
 
 t?ew Att 
 
 BW 
 
 u 
 
 ni/r, 
 
 . 
 
 
 
 . 
 
-69- 
 
 In general terms, the above interpretation is that the 
 orienftlted tube feet extend out in the direction of their orientation 
 and in this state are ready to give the step reflex upon contact stim- 
 ulation. In the absence of such contact stimulation there are reflex 
 connections between the myodermal sheath and the ambulaoral nervous 
 system of such a nature that the ray, by twisting or bending or both 
 sets itself more nearly at right angles to the actively oriented tube 
 feet. Fig. 18 illustrates the first movements of an animal inverted 
 during active locomotion toward a e. All of the extended tube feet 
 are protruded in the direction of the former anterior. Figs.. 19 aad 
 20 illustrate the movements of the arms as described already (p$ (, ) 
 wttch result in righting and in the ray assuming a position more nearly 
 at right angles with the oriented tube feet. During the righting pro- 
 cess the un stimulated tube feet remain extended toward the animal's 
 anterior. The rays a e. however, in accordance with the above princi- 
 ple, bend toward one another and down so that the tube feet come in 
 contact with the substrate, execute the step reflex and in the manner 
 outlined above initiate the righting. 
 
 The tube feet, however, have been regarded (Romanes and Swart 
 (1881), Preyer (1886), Loeb (1900), Jennings (1907), Moore (1910^1910^) 
 Cole (1913* )) as taking hold of the substrate and pulling the animal 
 over. Observation of the reaction as it occurs on sand show that this 
 pulling is not a fundamental or necessary part of righting. Pulling 
 by oriented tube feet is, however, a part of the step reflex. Since 
 attachment increases with the resistance to the step (pp-<lft-), and th 
 resistance to the step, in the initiation of the righting reaction, is 
 very great, it follows that attachment is tight and pulling is the most 
 noticable activity of the tube feet. It is this pulling, that has ob- 
 scured the eyes of observers, the more important and fundamental thing, 
 of which this pulling is merely a part, namely the step reflex. 
 

 cfi/J 5e. 
 
 . 
 
 
 
 
 . 
 
 
 ' - . . 
 
 
 fltj^UABP 
 
 ' fi.? t 
 
 . 
 
 
 
 
-69a- 
 
 If then the righting movements of the aims are dependent 
 upon the initial stages of the step reflex (oriented tube feet) 
 and the rioting movements of the tube feet are slightly modified 
 step reflexes, righting is itcelf a phase of loooraotion. 
 
 
 
 . - 
 
 T'LA^it 
 
 V-j i^Afd 
 
 
 t 
 
 'irVU, 
 
 

70- 
 from the faot that the stimulation o tha doreal 
 
 myodermal sheath of the ray JLj; not an essential factor In the 
 reaction* 
 
 If the righting reaction is simply a modification of ordin- 
 ary locomotion, it -would be expected first that contact stimulations 
 on the doraal myodarmal sheath of the ray do not play an essential 
 part in the ordinary locomotion and second, that, since the looomotor 
 impulse persists in a given direction for some length of time, the 
 righting reaction in the locomotor specimen shows a direction which 
 is very closely correlated with locomotion before and after righting. 
 Several laige active starfish were picked up when in rapid 
 locomotion and balanced inverted with the central part cf their disks 
 resting upon the bottom of a snail inverted beaker. Care was taken 
 in the manipulation to touch only trie disk and not to remove the 
 animals from the water or subject them to any other unnecessary stim- 
 ulation* In evary case two or more of the rays started to bend down 
 (dorsally) while the rays on the opposite side began to bend up* The 
 latter movements were more rapid than the former and the starfish coon 
 overbalanced and fell off the beaker* This was repeated so many tiates 
 that there is no doubt in my mind that the dorsoflexion and ventre- 
 flexion results of the operation of the "unified impulse" persisting / t 
 from the locomotion* That these movements are homologous with the 
 early righting movements (Jennings type 1) is indicated by the fact 
 that the rays which turn down turn also, usually, toward each other*# 
 
 # This conclusion is rendered more probable by the fact that 
 some of the neurotomized starfish when coordinated in locomotion 
 would show righting movements if inverted quickly and gently* These 
 movements were similar in direction to those of the normal animal 
 but were complicated by the fact that sooner or later these animals 
 tended to take the "tulip form". (Romanes and wart 1881). In a 
 few cases they righted themselves quite promptly. 
 
: 
 
 
 
 
 
 
 
 ' 
 
 
 pi' be. 
 
 
 
 
 
 
 
 M^MMMMMw*iW**M*MMMVW*<*> M '** | Mj 
 
 irivnfvR e 
 
 >LU9 
 
 ' 
 
-71- 
 
 Moore (1920) states that if suspended with the ventral side 
 down, an Aateriae forbesi -will remain motionless in a state of 
 ventral flexure indefinitely. This while not absolutely tru of an 
 active Pisaater especially at first, and very far from tjue of an 
 active Pyono podia, may be said to describe the behavior of th more 
 inactive specimens that I h>ve tried the experiment upon. Moore 
 says, furthermore, that if the dorsal wall of a ray of such a sus- 
 pended specimen be irritated by rubbing it with a glass rod, the 
 ray will flex dor sally. I have confirmed this. Moore, however, 
 neglects to mention a fact, first observed by Romanes and Swart (1881) 
 that the tube feet of such a ray whose dorsal dermis is irritated 
 increase in activity. The normal orientation of tube feet on an 
 active but unoriented speoiraan is toward the tip of the ray. It would 
 seem then that the dorsal flexure is due to the principle that a ray 
 tends to set itself more nearly at right angles to the actively or- 
 iented tube feet* This is perhaps the more acceptable as a point 
 of view since the activity of the tube feet has been observed to 
 spread to the tube feet of other rays and to be followed by dorsal 
 movements or lateral twistings of these other rays. 
 
 Moore comes to the conclusion from these and similar ex- 
 periments that the dorsal flexures of the rays which he has 
 
 ' 
 elicited by contact stimulations are the separate parts of the 
 
 righting reaction. Aside from the fact that the righting reaction 
 has been observed to start without any contact stimulation of the 
 rays, my observations and the statements available in the literature 
 have led me to the conclusion that lateral twistings of the rays 
 are muoh more important in the righting reaction (save that of 
 otsn) than are me^e dorsal flexures. 
 
 Evidence from the persistence ojf ffle ** uni f i ed iflipul s e " 
 

 
72- 
 
 It remains now to inquire into the correlation bet-veon the 
 direction of righting and that of locomotion before and after the 
 reaction* Sole (1913<) has presented some evidence on this point, from 
 which he draws negative conclusions* Hia analysis of the data is, I 
 think, incomplete and the data are not statistically representative* 
 
 He argues as follows* 
 
 "In table 4 are shown the results of a number of tests to 
 determine what relation exists betwean the arms used in righting when 
 the starfish is placed on its aboral surface and the direction of lo- 
 oomotion previous to and subsequent to thy righting reaction* The 
 data nay be summarized ap follows* 
 
 Arms e ed d de e ea a ab b JPQ 
 
 Crawling previous to test 2 6 5 1 3 2 
 
 used in righting S 2 16 1 2 
 
 crawling subsequently 2951 2 3 
 
 This shows that whereas the four spec i meats used in these taste 
 righted theasdlvae on arms a a sixteen out of twenty-four times, they 
 had been in nearly all cases crawling in a direction nearly opposed 
 to these arms, and mo reovor they continued locomotion in the same 
 general direction after righting themselves. An examination of the 
 individual records reveals the same relations in a great majority of 
 cages** 
 
 Itelow is table 4 to which column 2 and column 5 have been 
 added to help in interpreting the data* Cole's studies have led 
 him to the conclusion that the starfish studied crawls with anter- 
 ior, more than with any other rays anterior* Unfortunately, however, 
 in thasa experiments he chose animals that were not typical in this 
 respect, since in no trials were they crawling toward & and in all 
 but four trials were crawling in a vary different direction* Thit in 
 connection with the fact that only four specimens were used, all 
 presenting an unusual 
 
: : ^ &'i JS04X 
 
 ' 
 
 vo 
 
 * IK 30 
 
 
 
 ftv*Jr.3 . anri no 0rr leant 
 
 >,. tc it 001061 8 ni ^niv0*ro ae. - 
 oiU r.i nollQBOooX teuntfnoe v.orfj le 
 
 
 
 to* S omloe riciiiv o* * Xtf0r t 
 
 
 ^furrllib X"- 1 ' ^ ai lallvjrco *rer 
 
Table 4* 
 
 Relation of arms used in frighting to direction of 
 previous and subsequent crawling* 
 
 Individual Previously Anas used Shift of Subsequently 
 
 crawling in physiolo- crawled 
 
 ,4 After tri-il anterior righting anterior 
 
 "anterior" 
 (rays) 
 
 4 After trial 50, 
 
 d 
 
 ea 
 
 .0 efore " 1 
 
 - 
 
 ea 
 
 i After " 10,; 
 
 a 
 
 e(b) 
 
 ,0 After " i,}-. 
 
 
 
 i{ab) 
 
 ,0 After " 28^ 
 
 od 
 
 Ml 
 
 ,0 following day 
 
 - 
 
 Od 
 
 n ft 
 
 bo 
 
 bo 
 
 " " 
 
 a 
 
 od 
 
 .2 Trial 1 
 
 PI* 
 
 bo 
 
 .2 " 2 
 
 a 
 
 Hi 
 
 2, " 3 
 
 if 
 
 ea 
 
 ,2 " 4 
 
 od 
 
 a* 
 
 .2 " 5 
 
 e 
 
 M 
 
 .4*1 
 
 .. 
 
 
 
 A M 2 
 
 d 
 
 JU 
 
 .4 3 
 
 d 
 
 3l 
 
 .4 w 4 
 
 od 
 
 a 
 
 .4 5 
 
 od 
 
 M 
 
 .4 6 
 
 od 
 
 ea 
 
 .4 7 
 
 d 
 
 a 
 
 .4 8 
 
 d 
 
 ea 
 
 .4 9 
 
 od 
 
 od 
 
 . 10 
 
 
 
 ea 
 
 L4 11 
 
 bo 
 
 
 1.5 
 
 Sfcift of 
 
 anterior 
 from "pre- 
 viously 
 crawling" 
 (rays) 
 
 1.5 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 od 
 
 2.5 
 
 2 
 
 bo 
 
 .5 
 
 2 
 
 od 
 
 
 
 - 
 
 be 
 
 .. 
 
 
 
 a 
 
 1.5 
 
 2.5 
 
 ft d 
 
 2 
 
 
 
 a 
 
 
 
 .5 
 
 de 
 
 1.5 
 
 1 
 
 M 
 
 1 
 
 2 
 
 o 
 
 .5 
 
 2.5 
 
 od 
 
 .5 
 
 
 
 d 
 
 
 
 1.5 
 
 d 
 
 
 
 1.5 
 
 od 
 
 .5 
 
 % 
 
 od 
 
 
 
 2 
 
 od 
 
 
 
 2 
 
 d 
 
 .5 
 
 1.5 
 
 d 
 
 
 
 1.5 
 
 od 
 
 .5 
 
 
 
 stationary 
 
 > 
 
 
 
 bo 
 
 mm 
 
 .7 
 
 These trials were of a series of 490 showing the persistence of the 
 Biological anterior in a general direction, which tends to rotate 
 >wly to the ri^it or the left. 
 

-74* 
 
 direction of loco-action, Iea4a DM to believe that the data 
 not a good foundation for any fusion, Koreover the 
 
 tU*y do 
 conclusion Hr-doeo indicate is not that drawn by Cole. 
 
 As 0en from an exanimtirn of column 5, the 17 records show 
 that tJie Physiological anterior has shifted in one direction or the 
 other ^naveraee of seven tenths of an inter radius, per reaction* 
 Coles conclusion on this point, as seen above ie that "they^fbc *Ja *$'* 
 continued tc crawl, in the same general flireotion (an they did 
 Before) after righting themselves.* 
 
 :-3oYor f as seen from an examination of column 3, the 
 19 records show an averse athlft of anterior (referring to th 
 rays used to right as anterior) of 15 inter radii per reaction* 
 Coles conclusion on this point, however ia that the animals right 
 in a direction nearly opp+Bite to that in which they were pre- 
 viously (and subsequently) crawling. But the arithmetical 
 difference between theae averages of data (1*5 - 7* .8) is 
 8 of an interradius a shift which is approximately equal to 
 the shift (,7 interradius) which Cole regards ae no shift at 
 all. Obviously, then a detailed examination of Ooles data does 
 not ce-nfizta |ii conolueiona* 
 
 with an idea of clearing up the relationship between the 
 physiological anterior and the arms used In righting seventy* 
 five experiments were made with twenty-six tar fish (20 
 Pisaatar and 6 Aaterjna.)* The starfish used were in active 
 locomotion, except in case of some ef the Asterina as shown in 
 the record, manipulation WAS as gentle as possible, the animal 
 bein ? picked up by the disk and inverted qulokly without, in 
 meat eases, lifting it above the surface of the water* 
 Directive factors in the surroundings such .as light or areas of 
 shaddew etc*, were excluded by rotating the animal in successive 
 trials. 
 

 
 
 
 
 
Relation of arms ueed in righting* to direction 
 of previous and subeoqusnt 
 
 idividvnl. Direction Arms Shift Arms Arms Shift Direction Shift 
 
 before first of bent ri.jhted of after of 
 
 inrrtina^ bent anterior up on anterior rioting anterio 
 
 down, in radii ventrally. inradtt in radii 
 
 .wan 11234 567 8 
 
 >.l Hs 
 
 
 
 
 
 
 
 
 aster ,e 
 
 aed 
 
 0.0 
 
 bo 
 
 ed 
 
 1.0 
 
 3 
 
 0.5 
 
 2 ae 
 
 aeb 
 
 0.0 
 
 ed 
 
 e 
 
 0.5 
 
 ed 
 
 1.0 
 
 2 'io 
 
 aeb 
 
 0.0 
 
 bo 
 
 ab 
 
 1.0 
 
 ae 
 
 0.0 
 
 1 de 
 
 e 
 
 0,5 
 
 (a) bod 
 
 Ja)e 
 
 1.0 
 
 ae 
 
 1.0 
 
 1 e 
 
 o (e) 
 
 0.5 
 
 
 
 ae 
 
 0.5 
 
 ae 
 
 0.5 
 
 2 de 
 
 dea 
 
 0.0 
 
 bo 
 
 d 
 
 0.5 
 
 od 
 
 1.0 
 
 4 
 
 (i)b > jfi 
 
 0.0 
 
 bo 
 
 ... 
 
 .. 
 
 .. 
 
 .. 
 
 5 de 
 
 io(o) 
 
 0.0 
 
 bo 
 
 de 
 
 OP 
 
 ae 
 
 1.0 
 
 5 e 
 
 rjg (d) 
 
 0.5 
 
 bo 
 
 a (d)e 
 
 0.5 
 
 e 
 
 0.0 
 
 6 ?. 1 
 
 a(b}de 
 
 1.0 
 
 o(b) 
 
 de 
 
 1.0 
 
 d 
 
 1.5 
 
 5 o 
 
 bo 
 
 0.5 
 
 a 
 
 bo 
 
 0.5 
 
 bo 
 
 0.5 
 
 5 n b 
 
 bo 
 
 0.5 
 
 dea 
 
 be 
 
 0.5 
 
 
 
 1.0 
 
 5 " tf 
 
 eda 
 
 1.0 
 
 bc(d) 
 
 ae 
 
 0.0 
 
 ae 
 
 GL.O 
 
 7 e 
 
 1C 
 
 0.0 
 
 toe* 
 
 ea 
 
 0.5 
 
 ea 
 
 0.5 
 
 8 bo 
 
 (b)ctfd 
 
 0,0 
 
 (b)sit 
 
 d 
 
 1.0 
 
 cd 
 
 1.0 
 
 8 de 
 
 de 
 
 0.0 
 
 .. 
 
 de 
 
 0.0 
 
 d 
 
 0.5 
 
 7 od 
 
 b 
 
 1.5 
 
 a ode 
 
 b 
 
 1.5 
 
 b 
 
 1.5 
 
 5 de 
 
 de(o) 
 
 0.0 
 
 boa 
 
 M 
 
 0.0 
 
 de 
 
 0.0 
 
 7 ae 
 
 e(a)b 
 
 0.0 
 
 (b)od 
 
 ae 
 
 0.0 
 
 a 
 
 0.0 
 
 5 p 
 
 (b)ae 
 
 0.5 
 
 bo 
 
 ae 
 
 0.5 
 
 .. 
 
 - 
 
 9 ae 
 
 abod(e) 
 
 .. 
 
 e 
 
 ab 
 
 2,0 
 
 stopped 
 
 
 
 9 de 
 
 (u)do 
 
 1.0 
 
 ebo 
 
 d a 
 
 1.0 
 
 ae 
 
 1.0 
 
 10 * bo 
 
 bo(a) 
 
 0.0 
 
 V 
 
 ... 
 
 ... 
 
 -- 
 
 
 
 11 
 
 bo 
 
 be 
 
 0.0 
 
 ade 
 
 bo 
 
 OP 
 
 bo 
 
 00 
 
 1 7 
 
 Od 
 
 od 
 
 0.0 
 
 abe 
 
 od 
 
 0.0 
 
 .. 
 
 
 
 12 
 
 de 
 
 (a)ed 
 
 0.0 
 
 bo 
 
 ed 
 
 0.0 
 
 ed 
 
 0*0 
 
 13 
 
 
 (a bod) 
 
 0.0 
 
 (abode) 
 
 .. 
 
 - 
 
 * 
 
 .. 
 
 12 
 
 od 
 
 deo 
 
 1.0 
 
 ab 
 
 do 
 
 1.0 
 
 de 
 
 1.0 
 
 12 
 
 de 
 
 (a)de 
 
 0.0 
 
 bo 
 
 a ,> 
 
 1.0 
 
 e 
 
 0.5 
 
 12 
 
 de 
 
 e(a*)a 
 
 1.0 
 
 bod 
 
 M 
 
 1.0 
 
 jj. 
 
 1.0 
 
 12 
 
 2 . 
 
 ed 
 
 1.0 
 
 ab 
 
 ed 
 
 1.0 
 
 ed 
 
 1.0 
 
 14 
 
 
 
 da o 
 
 O.D 
 
 aed 
 
 do 
 
 0,5 
 
 e 
 
 0.0 
 
 15 
 
 4 
 
 (o) 
 
 0.0 
 
 tulip form 
 
 e* 
 
 -- 
 
 
 
 16 
 
 od 
 
 a 
 
 2.5 
 
 (e) (a)bo< 
 
 i if 
 
 2.0 
 
 Ml 
 
 2.0 
 
 16 
 
 ae 
 
 ae 
 
 0.0 
 
 bod 
 
 1,J 
 
 o.c 
 
 aa 
 
 0.0 
 
 16 
 
 If 
 
 ed 
 
 1.0 
 
 cba 
 
 ed 
 
 1.0 
 
 li'j 
 
 2.0 
 
 16 
 
 ab 
 
 ae 
 
 1.0 
 
 od(e)b 
 
 a4 
 
 1.0 
 
 
 1.0 
 
 16 
 
 ae 
 
 ao 
 
 0.0 
 
 bod 
 
 ae 
 
 0.0 
 
 ae 
 
 0.0 
 
 17 
 
 bo 
 
 bo 
 
 0.0 
 
 aed 
 
 bo 
 
 0.0 
 
 tto 
 
 0.0 
 
 17 
 
 bo 
 
 bo 
 
 0.0 
 
 aed 
 
 bo 
 
 0.0 
 
 bo 
 
 0.0 
 
 17 
 
 bo 
 
 bo 
 
 0.0 
 
 aed 
 
 bo 
 
 0.0 
 
 bo 
 
 0.0 
 
 18 
 
 od 
 
 bo 
 
 1.0 
 
 aed 
 
 bo 
 
 1.0 
 
 od 
 
 0.0 ^ 
 
 19, 
 19 
 
 d 
 
 J 
 
 abode 
 abode 
 
 - 
 
 (dea) 
 
 e 
 
 bo 
 bo 
 
 1.5 
 1.0 
 
 od 
 (b)od 
 
 ilo 
 
 19 Od 
 
 abode* (f) 
 
 .. 
 
 
 
 bo 
 
 1.0 
 
 
 
 0.5 
 

 
 
 
 . 
 
 . 
 
 
 )60 
 
 JMM 
 
 
46 No. 20 Aster- ae 
 ina 
 47 " 20 " ab 
 
 ae(b) 
 
 0.0 dob 
 1.0 abo 
 
 ae 
 
 TO 
 
 0.0 
 1.0 
 
 ^topped - 
 ab 0.0 
 
 48 " 20 
 
 n 
 
 de 
 
 (de)abo O.o ao(b) 
 
 de 
 
 0.0 
 
 ^e 0.0 
 
 49 " 20 
 
 * 
 
 ae 
 
 ae 
 
 0.0 bou 
 
 M 
 
 0.0 
 
 -"3 0.0 
 
 50 " 20 
 
 H 
 
 
 
 
 
 0*0 abod 
 
 o 
 
 0.0 
 
 stopped 
 
 51 " 20 
 
 i 
 
 ea 
 
 ea(bo) 
 
 0.0 dbo 
 
 a -i 
 
 0.0 
 
 ea 0.0 
 
 52 20 
 
 A 
 
 de 
 
 (d)ea 
 
 O.C bod 
 
 ea 
 
 1.0 
 
 od 1.0 
 
 Si " 20 
 
 H 
 
 
 
 o 
 
 0.0 tulip 
 
 form 
 
 mm 
 
 .. 
 
 54 " 20 
 
 14 
 
 d 
 
 d 
 
 0.0 (de) 
 
 de 
 
 0.5 
 
 stopped - 
 
 55 
 
 21 
 
 
 ae 
 
 ae 
 
 0,0 
 
 od 
 
 2.0 
 
 de 1.0 
 
 56 
 
 a 
 
 
 bo 
 
 be 
 
 0.0 aed 
 
 bo 
 
 0.0 
 
 stopped 
 
 57 
 
 21 
 
 
 da 
 
 abode 
 
 stopped 
 
 .. 
 
 mm 
 
 -- -- 
 
 58 
 
 23 
 
 
 se 
 
 ae 
 
 0.0 stopped -- 
 
 59 
 60 
 
 23 
 
 24 
 
 n 
 
 a a(e) 
 0(Bta)abode 
 
 0.0 bod 
 stopped 
 
 M 
 
 0.5 
 
 stepped 
 
 61 
 
 24 
 
 N 
 
 de 
 
 to 
 
 tulip fora 
 
 mm 
 
 
 
 .. 
 
 62 
 
 24 
 
 n 
 
 de 
 
 00 
 
 2.0 ade 
 
 bo 
 
 2.0 
 
 stopped 
 
 63 
 
 25 
 
 n 
 
 bo 
 
 bo 
 
 0.0 ade 
 
 bo 
 
 0.0 
 
 stopped 
 
 64 
 
 26 
 
 * 
 
 ab 
 
 Ml 
 
 1.0 bod 
 
 ae 
 
 1.0 
 
 * .- 
 
 65 
 
 25 
 
 * 
 
 Bo 
 
 bo 
 
 0.0 ade 
 
 bo 
 
 0.0 
 
 -- 
 
 66 
 
 25 
 
 H 
 
 ae 
 
 ae 
 
 0,0 bod 
 
 ae 
 
 0.0 
 
 mm 
 
 67 
 
 M 
 
 M 
 
 ab 
 
 MJ 
 
 1.0 bod 
 
 ae 
 
 1.0 
 
 mm 
 
 68 
 
 26 
 
 N 
 
 de 
 
 eaod 
 
 1.0 Tarn. 
 
 e 
 
 0,5 
 
 mm 
 
 69 
 
 23 
 
 H 
 
 3 
 
 de 
 
 0.5 dob 
 
 it 
 
 0.5 
 
 mm 
 
 70 
 
 22 
 
 N 
 
 de 
 
 de 
 
 0.0 oab 
 
 de 
 
 0.0 
 
 mm 
 
 71 
 
 23 
 
 M 
 
 ab 
 
 ae 
 
 1.0 odb 
 
 ae 
 
 1.0 
 
 * -- 
 
 Tl 
 
 25 
 
 Stationary 
 
 rigid 
 
 tulip form 
 
 73 
 
 24 
 
 * 
 
 n 
 
 
 
 * 
 
 * 
 
 * 
 
 
 74 
 
 25 
 
 H 
 
 * 
 
 1 
 
 
 
 M 
 
 * 
 
 
 75 n 24 
 
 H 
 
 N 
 
 
 
 M 
 
 * 
 
 I 
 
 
 36 
 (64 trials) 
 
 .6 
 (62 trials) 
 
 .57 
 (44 trials] 
 

 
 
-75- 
 
 Records were taken (column 1.) of the direction of locomotion 
 before righting and (eoluon 2) tha a IMS that, after in verting, 
 thf animal, first twisted and bent down tow-rd the substrate* 
 These two findings were compared in each experiment and the shift 
 in either direction of the leading rays or "physiological anterior" 
 set down in oolumn 3* The turning down of certain rays is 
 usually followed ( rscpac or preceded ) by a lifting up of others* 
 The rays that lifted up free of the substrate ** but not those 
 that were oriented on the substrate, in the way described ibova, 
 to walk over the initiating rays* were next recorded (column 4). 
 
 The ray a that turned down -were not, always, of course the 
 same is those that the animal uaea ia righting* ihese latter 
 are listed in ooluran 5, and the shift of anterior from the 
 direction bef re inverting to the araa used in righting is 
 listed in column six* The anterior after righting is listed in 
 column 7 and its shift from the direction before inverting is 
 listed in column 8. Blue the shifts of anterior, listed in 
 columns 3,6 and 8 refer to the original anterior before inverting* 
 
 A comparison of the averages obtained here* and those drawn 
 
 from Colt'a data shows that careful manipulation of the starfish 
 
 e. 
 and the use of a lirge number of individuals riduces the shift 
 
 of anterior considerably. .\s shown by the rays that are first 
 turned down, the anterior at the beginning of righting feas 
 shifted *38 of in inter-radius on an average of 64 observations* 
 As shown by the rays on which the animal ri t #its ) the anterior 
 during the righting reaction has shifted *6 of an inter-radius from 
 where it was before the animal was inverted. After righting, the 
 anterior shifts slightly back Coward its original direction, as 
 shown by the fact that the average shift after righting is less 
 than during righting* This shows more markedly in the average 
 

 e a&J 
 
-76- 
 
 drawn from Sole'w table. 
 
 This return of the anterior to'*ard its original direoticn 
 ia an example of the tendency which we lave noticed in connec- 
 tion with the deviation reaction ( p iff ) for the coordinated 
 impulse to return to its original direction, even after having 
 been actively oriented in aoue other direction* 
 
 le (1913) has shown very conclusively that the impulse 
 to locoraota, in the starfish tends Iks to maintain the same 
 general direction, from trial to trial. (Between each trial the 
 animal was held inverted by the disk until the raya dropped and 
 then "started" on the bottom of an aquarium in a non-directive 
 chamber.). Hie tendency to lK*ep in the sans direction was of 
 course only general, aa there wag also a rotation of the anterior 
 toward the rifc-Ut or toward the left, and certain aberrant 
 
 deviations, of from one half to two and a half inter-radii 
 
 i#u/n 
 
 Incurring quite frequently. In B eoanafrting up these deviations 
 
 Lu&) 
 
 from the table opposite It it was found that they amounted 
 to a sum total of 217 intor-rudii in 499 trials* 'Ihis amounts to 
 a shift of anterior of .43 inter-radii per trial which is quite 
 comparable quantitatively with the figures (,38 ,CO, .67/ inter- 
 radii ^obtained from the status of the direction of the coordinated 
 impulse ttiroughout the righting reaction* 
 
 Z conclude therefore that righting it an aspect of locomotion* 
 
I/ jriaas.tar ooraceus presents Uu three follo^in^ well 
 marked physiologic*! spates (1) "Rigid" (L) "loconiotor" (3) 
 "active but unorientod* i'he responses of th-i tutu feet, and 
 arms differ markedly according to tha physiological state of 
 the animal Other starfish studied present analogous states* 
 
 2/ intension of t ,e tuba feet depends upon the proper 
 physiological st^te and absence o< stimuli which cause Detraction. 
 
 An isolated tuba foot, infl-itod -*ith watar und ir pressure can 
 
 / 
 
 b3 caused to glonrly oxtend ; but not rjuite normally, 
 
 r? 
 
 3/ At^ching is condition-^ by the proper uTysiolo?ioal 
 
 state. An isolatyl tu :>B foot, properly prap^rud and i 
 with vfat^ 1 " is rnora apt to a -taoh if taken frota i ri^id starfish 
 than from a loooiaotor stirfisn. Attaching may involve only a 
 part of the ainbulaoral digk. 
 
 4/ .Uth'lr?nral A is a response to cont-iot stimulation, as is 
 dotaohing, under certain oon^ itions 
 
 I/ Th step '-eflex intargrades ^ith the withdrawal response 
 as elicited by 10 ivict stimulation of the ambulacral disk* It 
 is da ;'md3nt upon tirs contact stimulation uvl .t.^e presence of 
 
 the locoiootor irapulsefc vhioh orients the ?tep reflex and conditions 
 
 
 
 the tube foot to b-3 rigid and support animal during loootnotion* 
 The tu>>e foot is attached most strongly during the first part of 
 the step raflox* The tube foot is a tachad *ivh 2.8 (A.g farina.) 
 or ii,06 (ryono podia) times as much force as it exerts in pulling 
 against resistance* ITii^ facto/I- is relatively constant for vrious 
 values of the resistance. 'm& strorvjth of the step reflex 
 
 varies markedly with diff3;r3nt spscias* 
 
 /Joordin-tion of thj 'tuDa fjet of ti;e rigid starfish, like 
 
 tint of tha gills, is u siiaple apreid of exuension or rat^ction. 
 It ia ref rabla hypo the tioally to a simple rwrv3 rut. 
 
(*) 
 
 to . 
 
 
 
-79- 
 
 7/ Coordination in the active but unoriented starfish involves 
 orientation of tne distal tube feet, toward the tips of the 
 rays. With the rays on separate substrates, thi- -jncy 
 
 results in their -valuing in five dif fe^snt directions, ^nder 
 pathological conditions this ten enoy results in autotumy. 
 Orientation of the tuoe feet is not referable to a simple nerve 
 net as is Coordination in extension and -et-action but to a more 
 
 complicated and possibly an independent mechanism. 
 
 >L^ 
 
 8/ The unified impulse is formed (1) "by the spreading bink of 
 the oriented stale in tne tip of one of the r^ys. Various factors 
 may cause tne relative increase 'vhi.Cn results in its spread over 
 t.ie rast of t e animal (2) by the spreading baoV and fusion of 
 Reoriented states in adjacent rays. (3) By direct orientation 
 of tne tuba feet from exits, tion of the dermal nerve n;t or the 
 tube feat, themselves* 
 
 9/ .Behavior of the oriented animal is conditioned by all of 
 the above factors acting at the same time and in nice balance 
 against each other. In the actively migr*ting starfish the tube 
 feet are all oriented in tii3 same direction. 
 
 10/ The unified impulse^; (l) in some types of righting reaction, 
 
 (2) in the Deviation reaction, (3) in the looomotor starfish with 
 
 **s. 
 a curved lateral arm, is broken up into areas in Wjiioh the tube feet 
 
 are oriented in different directions. This is highly adaptive. A 
 possible -hysiolosical explanation is saen in th3 traction on the 
 tube feat resulting irom the movement of the rays over the sub- 
 strate. Evidence for tuis hypothesis is -rawn from (1) Ueurotoraized 
 starfish (2) starfish wit,, the rays placed on separate substrates; 
 
 (3) the mechanics of the deviation reaction. 
 
 ll/ The righting reaction is a phase of ordinary locomotion 
 

so 
 
 with the starfish in more or less a state of unifi-d coordination 
 The movements of the arms are explained on the assumption of re- 
 flex connections by '#hich the arms a^e bent or twisted more nearly 
 at right angles to the -actively oriented tube feet* Evidence for 
 this conclusion is n ra*?n (l) from the movements of tha tube feet 
 md arms: (2) from an analysis of tne reaction -*hen the tube feet 
 
 are proventsd .attaching by inv'irtin^ the Animal on sand; (3) 
 from the fact th >t stimulation of the dorsal myodermal sheath of 
 tn3 ray is not ->.n essential factor in the rig; ting reaction (4) 
 from the fact tnat the "unified impulse" persists during the ri ht- 
 ing reacti .n in Uie saiae direction to a decree quantitatively com- 
 parable to its persistence in ordinary locomotion (Cole). 
 

Baudelot, :ai. 
 
 x 
 
 1872. Ktudes general es sur la Systeme Nervaux. 
 
 Aroh. Zool. ixper. et Geti., 1, 177-216. 
 Bohn, G. 
 
 1908. Las asaais et lee erreurs cheat, lea etoiles de mer at 
 
 las ophiures. Bull Inst. Gen, Psych., 8, 21-102 , 
 
 T~ 
 
 56 fig*, in text. 
 Botazzi, F. 
 
 1898. Contributions to the physiology of unstriated muscu- 
 lar tissue. IV. Jour. Physiol., 22, 481-505, (p. 501 ff.), 
 
 ^-u- 
 
 22 figs, in text. 
 Clark, H. L. 
 
 1899. Synaptas of the New England Coast* Bull. U. 3. Fish. 
 Cornm., 19, 21-31, 2 pis. 
 
 **v*. 
 
 Cole, L. J. 
 
 1913a. Direction of locomotion in the starfish (Aeterias 
 
 fprbeaiJJ.iSxp. Zool., 14, 1-32, 5 tables, 9 figs, in text, 
 
 * -*-*t_ 
 
 19l3b. iSrperimants on coordination and righting in the star- 
 fish. Biol. Bull., 24^ 362-369, 2 figs, in text. 
 , R. P. 
 
 1909i. Preliminary Deport on the behavior of Schinoderms. 
 AJB. Kept* Director Dept, Marina Biol,, Year Book, 
 (Jarnegia Inst., vVaeh., 8, 128-129 
 
 i ^*^ 
 
 1909b. The movement of tne Starfish ^oninaater toward 
 
 the light. Zool. Anz., 35, 193-195, 1 fig. in text. 
 
 1911, Reactions tV light and other points in the behavior 
 of the starfish. Pap. Tortugas lAb. Carnegia Inst., 
 Wash., 3, 95-110, 6 figs, in text. 
 
1914. The difference of white and black walla on the 
 direction of locomotion in the starfish. Jour. 
 Anim. Behavior., 4, 380-382. 
 
 vK 
 
 Cuenot,L. 
 
 / 
 
 1888. Contribution a 1'iitude anatomique dee Aste'rides. 
 Aroh. Xool. Kxpe'r. et Gen ., Se'r. 2, 5, Supt., 
 mem II, 1-144, pi. 1-9. 
 De l^oore, J., and Chapeaux, M. 
 
 1891. Contributions a la physiologic nerveuse dee achin- 
 odarmes. Tidsohr. Jfedarl. Diark Ver., (2), Deel. 3, 
 1891, 108-169, pi. 7. 
 Driesoh, H. 
 
 1908. Science and philosophy of the organism. (Aberdeen 
 
 University) 2, XVI + 381 pp. 
 
 />\- 
 
 Graber, V. 
 
 1885. lieber die Helligkeits - und Parbenempfindlichkit 
 einiger Meerthiere, S.B, Akad Wiss* Wien F.ath. 
 Hat. n.,91,abt. I, 129-150, 35 tables. 
 Grave, C. 
 
 1900. Ophiura brevispina. Mem. Hat. Aoad, Sol., 3, 77-100 
 pi. 1-3, 5 figs, in text .(F Mem. 3iol. lab. Johns 
 Hopkins Univ.,^No. 5. 
 Greenwood, li. 
 
 1890* On the action of nicotine upon certain invertebrates 
 
 J. . Physiol., 11, 570-605, 
 Hess, C. 
 
 1914. Unterzuohung ufber dem Lichtzinn bei Jcninodermen. 
 
 Pflu^ger'a Arch., 160, 1-26, 6 figs, in text. 
 Holmes, 3 . J. 
 
 1911, The evolution of animal intelligence. (New York, Holt), 
 V<- 296 pp., 18 figs, in text. 
 

 
 
Jennings, H. S. 
 
 1907. Behavior of the starfish Aetarias aertulifara 
 
 Xantua* (A. forreri Oe Loriol). Univ. Calif, 
 
 Publ, Zool. 4, 53-185, figs, in text. 
 Krukenberg, 0* Fr W 
 
 1881. Betrage zu einer Narvanphyaiologie der ^Quinodoraen. 
 
 In Verg. Phya. Studion (Heidelberg, Tarl Winter's 
 
 Buohverhdg. ) 2 Reihe. 1 abth, 76-82, 2 figs, in 
 
 text. 
 Loeb, J. 
 
 1900. Comparative pliyaiology of the brain and comparative 
 
 psychology, (New York, Putnaa), X 309 pp., 39 figs. 
 
 in toxt. 
 
 !,., ynd liamam, 0. 
 1899, Seesterne in Bronn'a Klasaen and Ordungen da Thier- 
 
 Reiohs. (Leipzig, C* P. Winter* *che Buohverhandlung) 
 
 2. Abt. 3, 461-744, pla. 1-12, 13 figs, in text. 
 , S* 
 1908* Studien der I'hyisiologie des WerveneyatonB der JSohino- 
 
 dertnen, X. Die Ifagsehen der Beentdrne und die 
 
 Koordination ihrer Bewegun^en, Pfluger's Archir. , 122, 
 
 318-360, 14 figa. in text. 
 ItfOSb Studien d0r Phyeioldgie dea Hervenyatem der /.ohino- 
 
 darmen, II, Ueber daa Hervenoystem der Seeetarne und 
 
 uber < T .en Tonus. Pflugere Arohiv. 123, 1-39, d figa. 
 
 W^ 
 
 in text. 
 
 1909a. Sinne* phyciologie Studion an :,ohinodeTtr<jn. Zeit. 
 f. Allg. Physiiol,, 9, 112-140. 
 
 V-V. 
 
 1909b. Studien zur Phyoiologie dea Uerverisyc tarns der '^ohino- 
 d&rmen III Ueber die Armbewegung'^n dsr Schlangensterne 
 und. V* Hexkull'a IFundsuaentalgaaetz fur den . 
 

 , 
 
 
 
 
 
 ' 
 
verlauf, Pflugers Arcniv., 126^ 371-406, 6 figs. 
 
 in text and 4 tables. 
 llaet, R. 0. 
 
 1911. Light and the behavior of organisms. (New York, 
 
 Wiley) XI -I- 410 pp., 3? figs, in text, 
 iieyer, R. 
 
 1906, Unt : rsuchung uber <?.en feineren Bau des Nerven systems 
 
 t c r / s: t <: r i <? en ( AgteriaE rubens ) . 
 
 Zeit. Vice. Zool., tt, 96-144, plr. 9-1C. 
 Moore, A. R. 
 
 1910a. On the riffrtinf; movements of the starfish. Biol. 
 
 .11. 14, 235-239, 1 fie. in text. 
 l?10b. On the nervous mechanism of tha righting movemente of 
 
 the starfish. Araer. J. Physiol., 27. 207-211, 2 
 
 fign. in text. 
 1916. The action of strychnine on certain invertebrates 
 
 J. Phnn/i and Exp. Therap., 9, 167-169. 
 .\917. di3,acal difforontiatior: in the nervous syetem of 
 
 invertebrates. Proc. Nat. Aoart. Sol, 3, 59B-602. 
 1918, ^Hav^r^al of reactions by means of strychnine in 
 
 plans riana ^nd starfish. J, Gen. Physiol. ^, 97-100, 
 
 3 fi e i a. in text. 
 1920a The action of atryohnine and nicotine on the neuro- 
 
 musonlar raschaniscj of Asjteria^s. J. Gen. Phyaiol., 2, 
 
 201-204, 5 figs, in text, 
 1920b Steraotropism as a function of neuromuscular 
 
 organization. , J. Gan. Phyaiol., 2, 319-324, 4 
 
 *VA^- 
 
 figs* in text. 
 Hori-oan, '.".', W, 
 
 1900. Do the reactions of the lovor -aniraalo against injury 
 indicate pain senaationn. Amor, J. Physiol., 3, 
 271-284. 
 
Parker, G. H. 
 
 1919. The Elementary Nervous System. (Philadelphia 
 
 Lippincott) 229 pp., 53 figs, in text. 
 Plessner, H. 
 
 1913. Untersuohen iiber den Physiology der Seesterne. 
 Zool. Jarb. Abt, f. Zool.,33, 361-386, 7 figs, 
 
 ' AM- 
 
 in text. 
 Preyer, W. 
 
 1886. Ueber die Bewegungen der Seesteme. Mit. a.d. 
 
 Zool. Stat. zu. Heap., 7, 27-127; 191-233, pi. 7, 
 
 >*^-i 
 
 27 figa. in text. 
 Prouho, H. 
 
 1890. Du Sens de 1'Odorat ohez. les e'toiles de msr. 
 Gompt. Rend. Aoad. Sciences, Paris., 110, 
 
 A-*-^- C F 
 
 1343-1346. 
 Reamur, R. A. P. de 
 
 1710, Du mouvement progressive et de quelqu^s autres 
 
 mouvements de di verses espeoes de coquillages, 
 
 orties, et ^toiles de mer. Mem. Acad. Roy. 
 
 Science Paris, 17^0, 439-490, PI. 9-12. 
 Romanes, G. J. and Kwart J. G. 
 
 1831. Observations on the locomotor system of the 
 
 jjJohinodermata. Philos. Trans. Roy. Soc. Lon^. 
 
 3^72, Part HI, 829-882, pi. 79-85. 
 Romanes, G* J. 
 
 1883. Observations on the physiology of the eohinodermata 
 
 J. Linn. Soo. 17, 131-137. 
 1885. Jellyfish, starfish, and sea-urchin. (New York. 
 
 Appleton)XI t 323 pp, 63 figs, in text. 
 

 . 
 
 
 rajs ?b ee *I .xf-rfr jf nl 
 
 
 eb % . f UBM*A 
 
 ? OD Je 0T 
 
 ,5/i.-- : , t :^ sc- 
 
 . .ir^l ec reel 68 
 
 . 
 
 
Rusao, G. 
 
 1913. Analisi e mecoanismo del reflesso di raddrizzamento e 
 di altri raovimenti ooordinati negli eohinodarmi. 
 Atti. aco. Gioania Catania, Serie 5, 6 Mem. 22, 1-14, 
 
 Seheinmetz, P 
 
 1896. How do starfish open Oysters J. Jiar* Biol. Ass.. 4, 
 
 / *" 
 
 n.s., 266-268, 9 figs, in text. 
 Steiner 
 
 1898. Die funktionen des Cent talnervenays terns und if? re 
 Phylogenese 3, Die tfirbellosen This re, (Braun- 
 schweig Priederioh Vieweg) X * 154 pp., 1 pi., 
 46 figs, in text. 
 Sterne, 0, 
 
 1891. Five souls with but a single thought, the 
 
 psychological life of the starfish, luonist, 1 
 
 M- 
 
 245-262, 6 figs, in text. 
 Spix. 
 
 1809* Memoire pour servir a I'historire de l*Asterie 
 
 3 
 rouge, etc. Ann. Mus. Hist. Nat., Paris, 438- 
 
 A 
 
 458 pi. 32-33. 
 
 Tiedemann, I< . 
 
 1815. Beobachtung liber das Nervonsystem und des sensibilen 
 Brscheinungen der Seestame.Meokal's Arch.f. d. 
 Physiol., 1, 161-175. 
 
 4V 
 
 Verrill, A. S. 
 
 1914. Monograph of the shallow water starfishes of the 
 north Pacific coast from the Arctic Ocean to Calif- 
 ornia. Harriman Alaska Series of the Smithsonian 
 Institution Publications. 14, pl.l., XII <- 408 pp. 
 

 - 
 
 . 
 
 
Von Uexkuell, J. 
 
 1900. Die Physiologic des Seeigelatachela. Zeit. f. 
 
 Biol., 39, 73-112, 4 figa. in text. 
 1903. Studien ueber den tonus I. Der biologeaohe 
 
 Bauplan von Sipunculua nudus . Seit f, Biol., 4Jt, 
 
 269-344, pi. 6. 28 fige. in text. 
 Vulpian, A. 
 
 1866. Lecons sur la physiologie generale et oompare'e du 
 
 syatame nerveux, faites au Museum d'hiatoire naturelle 
 
 (Paris. Germer Bailliere) 737-742. 
 


 NON-CIRCULATING BOOK 
 
 72H925 
 
 UNIVERSITY OF CALIFORNIA LIBRARY